Full text of "Bob_Pease_Lab_Notes_Part_6"

See other formats

A Tribute to Bob Pease

Pease Porridge — 6 (2005-2007)

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHATS ALL THIS RESONANCE
STUFF, ANYHOW?

I was walking down a hallway on the top floor of NSC's building D, when something I heard made me suspicious. It sounded like a series of
tiny clicks rattling around the hallway. I stopped and slapped my thigh to make a sharp sound. I heard TICK-TICK, tick-tick, tick-tick. I was
surprised, because there was no obvious reason why this hallway should do this. It even had a carpeted floor. I tried the slap in several
other hallways—and in tile bathrooms—and got almost none of this effect. What's going on?

I decided to tape the effect, and if I could see the timing of the clicks on my scope, that might be a clue. After all, the hallway goes 270 feet
from one end of the building to the other. I'm standing near the middle when I get the best reflections. But there are other frequencies and
time relationships in there to study. I'll let you know what I figure out.

I was lying on my bed a while ago, humming, when I heard a strange humming below me. I tuned my hum up and down—there it was at
336 Hz. The mattress was resonating. The funny thing is that this is not just a simple resonance or simple forced damped harmonic motion.
When I stop my burst of humming, the amplitude then grows for another half-second, and then it slacks, and then it grows even bigger.

Let's see if I can tape this. I may need a limiter amplifier so the recorder isn't fooled by its AVC loops and the initial hum doesn't drown out
the rest. See the oscilloscope photo for the results of my taping experiment.

Nope, sorry. Even with the preamp, I couldn't get a clean recording. So the scope photo shown is just a (poorly) simulated waveform.

I've been on the road for quite a few days this fall, and every time I come to a new hotel, I hum at the mattress. I hear a little resonance, but
not nearly so much. I haven't got any good calibration for all those frequencies. I'm not going to say that my mattress at home is all worn
out or is of lousy construction. I'm just saying that if I had to invent a circuit or structure to make the resonance act like that, I would be
nuzzled how to do it. I onlv hear one freauencv. not two.

When I go hiking up in the Marin Headlands, or Fort Funston, there are several large concrete tunnels left over from World War II. They're
perhaps 20 feet high and 200 feet long. My sons early on figured they could hum and set up some great resonances. I don't know what the
Q is, but—several seconds. A Q of hundreds... Of course, they called this "Tunnel Hum." It must drive the other tourists and hikers crazy.

We can't resist it!

Many barbershop quartets like to rehearse and record in a concrete stairwell, at least for some songs. They really seem to like that
resonance. With only four people in the stairwell, they don't damp out the resonance as they would if there were 40 people. But stairwells
seem to provide better results even than a tile bathroom.

I go to church at Grace Cathedral in San Francisco. This is, structurally, a great beautiful concrete barn, and it has a seven-second
resonance. This means the organist has to wait after a loud chord and wait a couple seconds more than usual before playing the next chord.
When they get the church more than half full, this cuts down to perhaps four seconds. Brass players specifically have to wait for the bright
tones to die out. It makes a difference how you play!

What resonant frequency and Q do you get out of your mattress? I gotta slide over to my wife's side and see if its frequency is any different.
That might explain why the energy comes up gradually. But the Q has to be well over 100.

Comments invited!

GRACE CATHEDRAL IN SAN
FRANCISCO. THIS IS,
STRUCTURALLY, A GREAT
BEAUTIFUL CONCRETE BARN,
AND IT HAS A SEVEN-SECOND
RESONANCE.

January 13, 2005

Electronic Design

• •

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB’S MAILBOX

Hi, Bob: I have a problem and could use your help. It initially sounded easy to me, but proved not so. I need to find an analog circuit for
the following problem: I have 13 resistors with different values. I know the list of values and need to randomly pick two resistors from
those 13 to put into the circuit.

The circuit should be able to tell me which one is the larger resistance value. If possible, please provide some advice on this problem that
has puzzled me for a while.

Tiejun (via e-mail)

Pease: Hello, Tiejun. It might be of interest for you to study the range of those resistors. If you load a 0.1- and a o. 2 ->^' resistor onto a
5-V bus, you might get a stupid answer if they are destroyed by overheating. There are many ways to compare resistors' ratios. Some use
an op amp, while others use a comparator. Most use a matched pair of resistors as a divider, so if the unknowns are close together in
ratio, you won't befooled. This is usually called a bridge, such as a Wheatstone Bridge. Did they ever teach you about bridges, which
were used for 110 of the last 120 years? Set up your unknown resistors as a voltage divider. Set up the matched resistors as a voltage
divider. Provide a bias voltage. If the resistors are ill-matched, it is very easy to see: The comparator or op amp tells you which R is
HIGH.

Dear Bob: I read your Pease Porridge column all the time and love it. Now I have a question for you. I have a component that I have to
source out and I'm having trouble finding it. The logo on the part is of a pair of hands doing a handshake. Do you know if this is an old
National Semiconductor symbol? (I sure doubt it. /rap) The part is some type of video amp for a CCD array. Any info would be greatly
appreciated.

Steve A. (via e-mail)

Pease: Man, I never heard of this nor saw this. I'll ask around. And perhaps some of our readers might know?

Dear Bob: Regarding "What's All This Magazine Stuff, Anyhow?" (electronic design, Nov. 15,2004, p. 18). The word "magazine" is
derived from the Arabic word makhzan, meaning storehouse per www.bhag.net/word/word24aug2003maga.html and my memory as
someone (with some Arabic) who lived in Morocco for two years.

Fred White (via e-mail)

Pease: Hello, Fred, I am a bum for neglecting to include the etymology. I always like to do that, and I can't think of why I neglected to
include it in my column. Thank you for mentioning this. But if you went back to Morocco, what do you think a storehouse is now called?
If I went back 500 years to old England, what do you think a storehouse was called?

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

TREKKING WITH BOB PEASE

Anybody interested in a great bicycle trek around the Annapurna Circuit? It was pretty tough when we did it in 2002, but I got a couple
more friends who want to do it, so I might just go again. We'll be starting up the trail on June 1, 2006. Read about that trek at
www.national.com/rap/trekking/bek/index.html, parts 14 and 15. Only tough, strong, mountain bikers AND good hikers should apply.

But it will be a lot easier for us than our 2002 trip, as we know when and where to hire porters, and how to keep our weight down. Inquire
to rap@galaxy.nsc.com.

Is anybody interested in joining us on our next hiking trek in November 2005? This is an easier hike, up from Pokhara to the Annapurna
Sanctuary—just 14,200 ft, with the great Annapurna peaks rising 10,000 ft above us, all around, and the full moon overhead on November
15. Our highest camp is at 12,200 ft. Then we traverse to Ghorapaani and Poon Hill for more great views. This is just a 25-day trek. Pricing
is around $1600 plus -$1400 for airfare from the west coast. For more information, go to www.instantweb.com/p/peter owens
/Annapurna%20Sanctuary%20Camping2004.htm. Save up for your vacation now. Inquire to rap. Do it early because space is limited to
about 12 people, /rap

January 20, 2005 Electronic Design

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHAT’S ALL THIS COMPARATOR
STUFF, ANYHOW?

There are many comparators that you can buy to provide quick (sub-microsecond) response when a large signal changes and crosses a
threshold voltage, such as a reference voltage. Unfortunately, comparators don't work well when the input signals are very small. The
ability to respond correctly, without offset, drift, or noise, is normally impossible to do with a comparator unless the signal is moving more
than a millivolt beyond the reference voltage. And with a comparator, you can't add a chopper-stabilizer because the comparator's offset
voltage is unknowable when the input signal applied to the inputs may be many millivolts or volts.

Here's an application where the LMP2011 can act as a precision comparator with better than 10 pV of resolution and precision. Yet by
closing its feedback loop continuously, the op amp maintains full dc precision, low offset, and negligible drift.

In Figure 1, the signal is brought in to the summing point through Ri. If the desired threshold point is zero, R2 isn't needed. (But if a
particular reference voltage Vref is needed, Ri and R2 must be closely matched so the output will trip when +Vin crosses (Vref).)

Alternatively, if the reference voltage is small (less than 100 mV) and at low impedance, it could be connected to the positive input of the
LMP2011 through R4.

The summing point voltage of the LMP2011 is maintained within a few microvolts of its positive input, basically all the time. When the
input crosses zero millivolts in a positive direction, the output starts to move down from its limit value, such as +1.4 V. Positive feedback is
applied through the R5-R4 divider to drive the positive input negative, increase the response speed, and supply dc hysteresis. Further, ac
hysteresis is applied through the R6-R4 divider. So for a short time, additional positive feedback voltage is applied to the positive
input—but only for a short time. When R6-C1 has discharged, the hysteresis will be restored to the dc value of about 15 pV p-p, or ±7.5 pV.
This is important to get freedom from noise or oscillation when the input signals are very small.

Using 2N3904S as diodes is very important for full accuracy because most ordinary diodes are much too leaky around ±60 mV to work
well. Ordinary gold-doped 1N914S or 1N4148S are quite unsuitable due to their high leakage and conductance, even at room temperature.

The observed delay time for signals as small as ±10 pV is about 5 ms, decreasing to 0.8 ps for large signals. See the chart of delay time
versus signal amplitude . In some cases with small input signals, less than 1 mV, the delay time can be reduced by a factor of three or four
by using lower values for Ri and/or R6, such as 2 k^. The improvement in delay time isn't quite as big a factor for large signals. Using low
values for R6 may cause dynamic errors or delays if the input waveform is asymmetrical in amplitude or in duty-cycle.

Comments invited!

VW R\, 5*

NM/*

C-Vuif)^

/ ki

p.6,Z°K • dtl

f,yr< t

February 03, 2005 Electronic Design

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB’S MAILBOX

Hello Bob: I am looking for professional opinions/experiences regarding pc-board assembly using eyelets. I was told by Sam Reaves that
you had written on the subject and could probably offer some good information. The problem that I am facing is that we are using eyelets
on some pc boards in an effort to make the joints stronger. But I have seen some failures that are due to the eyelets. I need information to
lead me to the correct decision and to use to convince others.

(Yeah, in my pretty good book on Troubleshooting Analog Stuff (p. 55), I said about 14 years ago that many people had given up on
eyelets. They are unreliable when cycled over many cycles. Myself I preferred to use double-plated-through holes, and I put wires
through, and nobody ever argued with me that my solution was wrong, compared to eyelets, /rap)

I am of the opinion that the eyelets do not offer much insurance for a good solder joint, and in some instances they can and have become a
liability. (You have shown that eyelets have lousy reliability. [That is a technical term.] It is not to put the onus on you to show that the
eyelets are lousy. If the guys selling eyelets cannot show a mode of applying the eyelets to make them reliable, then throw them out in two
minutes, /rap) What is your experience regarding eyelets and solder-joint quality? (I have had very little experience myself, but the guys
who say they are bad have never been, to my experience, contradicted. The burden is on the peddlers, and I'd be surprised if they can beat
it. /rap)

The application is an electronic lighting ballast. The pc board is single-sided, and the eyelets are used on the BJTs. I prefer plated-through
holes, but because of cost they are not very practical. The eyelets were added to the BJTs because it was believed that they were one of the
highest-stressed joints on the board, both electrically and mechanically in the manufacturing process. Because of the inconsistency I see
with the eyelets, I think I may be better off with a regular nonplated-through-hole pc board. What is your thought on that?

Jason Cook

(via e-mail)

Pease: As you have a pc board where plated-through holes are not cost-effective, can you add wire stitches to connect through the
board? Try some experiments. Add bigger foil lands for that transistor to bear the stresses and get the heat out. Your pc-board vendor
must have some advice.

Dear Bob: In your recent "Bob's Mailbox" (electronic design, Nov. 29,2004, p. 20), you mention that the low-leakage diode trick with the
2N3904 (my favorite part) should leave the emitter unconnected. (The column I just wrote for the last issue uses 2N3904S, too [electronic
design, Feb. 3, p. 18]! /rap) Why can't the base and emitter be connected together and still yield the low-leakage characteristic of the
base-emitter junction?

Dave B.

(via e-mail)

Pease: It will do no good and no harm to connect the B to the E. But it is important not to connect the base to the collector. That makes
an excellent diode, very fast (sub-nanosecond) with low leakage—but just for 3 or 4 V.

Hi, Bob: Your column on "Merit Badge Stuff' (electronic design, Dec. 18,2004, p. 20) inspired me to e-mail you. I earned the Electronics
merit badge in the late 1970s. And the Computers merit badge, too. Funny that I ended up majoring in biochemistry for my first semester
of college and then switched to EE. Been a professional EE for over 16 years now with only one month out of work last year.

Now I think my nine-month-old son will grow up to be an EE, too. I just found out that he was born on Gustav Robert Kirchoff s birthday
(March 24th)! Isn't this a great way to make a living? (Pointy-haired bosses notwithstanding.)

Jim Ford (via e-mail)

Pease: There are many ways to have a satisfying job in electronics! Even if you were—or weren't—born on Kirchoffs birthday. Best
regards, and best wishes to your son.

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

February 17, 2005 Electronic Design

• •

J PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHAT’S ALL THIS
STUFF, ANYHOW?

Recently I got an e-mail from a friend, bemoaning that he ran out of breath while trying to climb a 14-er in Colorado. He
was upset that he had to turn around at the 11,000-ft level and go back down with the women. I suggested that he use
the Bob Pease Conditioning Scheme. You don't have to drag a "Stair-Climber" or treadmill inside a simulated high-
altitude chamber, evacuated to 14,000 ft. You just have to walk and run up and down stairs—and not in a 20-story
building (though I've done that, too). I do this running in my own house, where one flight of stairs is 9.90 ft high. I step
with both feet on the ground floor and with both feet on the top level. I keep a good count. I keep up a good pace for 20
minutes. Then I lie down to rest a few minutes, and cool off, and take a shower.

I start out fast, two steps at a time ascending, and I scamper down. I slow down gradually to four or three flights per
minute, and I always breathe through my nose. That's a good way to get your lungs in shape for high altitudes. Running
up is great for your uphill muscles, and running down is great for your downhill muscles. Back in 2000 ,1 got in really
good shape for trekking in Nepal. I ran up (and down) 33,000 ft of stairs over several months. Then when we hiked up
to Everest Base Camp, at 17,800 ft, and Kala Pattar at 18,600 ft, I was definitely in the best shape of all the (younger)
guys. Muscles breathing. I walked away from them. Of course, you have to keep hiking (or skiing, or bicycling) or
whatever you are going to be doing. You need that, too.

How do I count the number of flights? I don't literally them. I time them. I use an analog clock with a sweep-second
hand to note what rate I am making, such as four flights per minute, exactly on the quarter minute. If I've done two
minutes at five flights per minute, (every 12 seconds) and six minutes at four per, and then nine minutes at three per,
plus a couple flights that I "gained" when I was going a little faster than three per, that adds up to 63 in 20 minutes. It's
a surprisingly good way to "count." And surprisingly good exercise. And a well-calibrated method. It's easy to not cheat!

At the foot of the stairs, along with the clock, I set a big glass of water, which I need at -12 minutes, and a towel to dry
off my face, etc. When I'm in good shape I can do 80 to 85 flights in 20 minutes, in the comfort of my home, any hour of
day or night, rain or shine. (Or in a hotel.) Usually, I use the stair-rail, but sometimes I don't. Obviously, I have to be
careful not to trip or fall. Sometimes I wear light shoes; sometimes my heavy boots. I'll be doing a lot more of this
shortly to get in shape for my next treks electronic design, and for some snowshoeing.

DISCLAIMER

Any exercise program must be taken in moderation, in view of your own physical condition. Don't try to start too fast.
Start at an easy rate, and work your way up to a vigorous rate. It's a good idea to keep your heart rate below your
recommended rate for your age. Some people say 200 minus your age is a good maximum heart rate. If you are in doubt
about your health for such vigorous exercise, consult your physician. Be careful not to fall on the stairs. And if any part
of you gets unhappy with the exercise, slow down and take it easy!

And, for best acclimatization, try to ascend no more than 1000 ft/day, as we do when we're trekking in Nepal. In
Colorado, nobody does that!

CONDITIONING

Comments invited! rap@galaxy.nsc.com —or:
Mail Stop D2597A, National Semiconductor
P.O. Box 58090 f Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

March 03, 2005

Electronic Design

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHATS ALL THIS POWER
STUFF, ANYHOW?

Sometimes* a tot is just right. Sometimes 200 hp in a

car is a lot better than 100. Sometimes even 1 hp gets a bicycle
up the hill quite fast. Yet a bicycle can cruise along at 6 or 8
mph with less than 0.1 hp. That’s part of the elegance of the
bicycle, as an extremely efficient mode of transportation.

Yet by using external sources of power, I can go uphill much
faster than I can by pedaling my bike. The dollars per mile
can drop way down, if I count my time at even $5 an hour.

A 747 at takeoff is generating perhaps 50,000 hp. Yet
its controls rely on tiny forces and low-power circuits, in the
milliwatt or microwatt levels. Radio signals may be kilo¬
watts when transmitted, but nanowatts when received. We
engineers require the ability to comprehend this entire
range of power—and to engineer control circuits that are
appropriate.

Power is useful when it’s in the right place, yet it can be
disastrous in the wrong place. So, our task is to keep power
in the right place. Sometimes a plain old linear regulator
with an efficiency of 30% to 50% is just right. Many other
cases definitely call for a switch-mode regulator.

Recently we had a high-power, high-current, high-speed,
medium-voltage driver. A potential customer asked us if we
could soup it up from 5 V of V s to 8 V to drive a couple hun¬
dred milliamperes into a higher-voltage load. Well, we said,
it does no harm to try. We applied the juice and monitored
the temperature of the calibrated ESD diode, which was at
the edge of the die. It went up to 175°C. One of the engi¬
neers said, “Well, it’s a good thing the temperature is con¬
stant, all across the die, because we really don’t want it to
get hotter than that, for best reliability.”

But a couple of us guys were suspicious. I computed that
the middle of the die could be 60°C or 70°C hotter than
the edge. We got mad and etched away the encapsulation
and took an IR photo. The middle of the die was indeed
85°C hotter than the ESD diode. It ran at 260°C, all week¬
end. On Monday, it was still working fine. Then we had to
tear down that experiment to do other tests. But we showed
that this part could handle a lot of power, get the heat out,
and keep running well.

We wouldn’t want to recommend that the part would run
this hot, for good reliability. But if we added a decent
heatsink, we could keep it cool enough to run reliably. Get¬
ting the milliamperes out does make it challenging to get
the heat out, but not impossible. (A metal heatsink does a
much better job than a paper heatsink.)

This computer-simulated photo shows how the heat is con¬
centrated in the center of a hot die.

808 f S M AIL B OX

Dear Bob: About “Resonance Stuff” (electronic design, Jan.
13, p. 22): That is an interesting tale and one that I have also
experienced! My wife inherited this cabin up in western Mass.,
and it was a very basic structure with a simple bed that had a
coil spring frame to support the mattress. The open springs
had very high Qs, so any humming or spoken sound while lying
on the bed was amplified considerably! I never took the trouble
to haul up a mic, preamp, and scope to analyze this, but we
used to joke about the springs doing this! I don’t quite follow
what the picture of the scope trace in your article is displaying,
other than that it looks like bursts of some frequencies!

Dale B. Blackwell (via e-mail)

- Pease: As I said, it was a poor simulation. The real ampli¬
tude (as soon as I stopped humming) was at a moderately
low level, and then the amplitude rose by 3 dB, then fell by 3
dB , and then rose 6 dB, then fell off, after a total delay of a
couple of seconds. OO

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

. 20 *

Gotowww.elecdesign.com 03.17.05 Electronic Design

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

ALL WHAT’S MAILBOX THIS
ANYHOW, STUFF?

Dear Bob: Regarding Jason Cook’s question on eyelets/ 1 I worked
at Motorola in the ’60s, when eyeleted pc boards were still
being used. In the days before plated holes were common, eye¬
lets did prevent foil from peeling when a board was repaired,
and they were the only way to get vias on two-sided boards.

The eyelets were specially designed "double funnel" and had
to be inserted properly so the funnel would split on the top side
to allow solder to wick up.

But they did cause headaches because of solderability prob¬
lems. We found that the plating had a shelf life. If the eyelets
weren't used promptly, the surface would corrode or get con¬
taminated. Or if they rattled around too long in the eyeleting
machine’s hopper, the plating would wear off. A bad eyelet was
impossible to see and very difficult to repair. Not only that, a
bad eyelet would cause an intermittent fault that might not
show up until the board was temperature-cycled or shipped to
a customer. As soon as plated holes became affordable, the
eyelets were all designed out.

♦ Ken Lundgren (via e-mail)

• Pease: I always heard that the eyelets would adhere to some
of the nearby foil, and then when thermal expansion
occurred, the eyelet would lift this foil away from the main
part of the foil, and the foil would crack in pieces. Maybe
that's not the major fail mode, these days. I have used thin
wire through holes and then soldered over, top and bottom.
Any two of those, in parallel, are much better than an eyelet,
in my opinion. (We have done this with homemade boards,
when the delivery on plated-through holes was too slow.)

Dear Bob: Your column on “Resonance Stuff" (electronic
design, Jan. 13, p. 22) was especially noticed by me because of
my own experience in my childhood. In 1945, at the tender age
of 12,1 developed an interest in ham radio. I thought I would
build myself a crystal radio using a galena crystal. Nobody
seems to know what that is these days, but you surely do!
(Galena is lead sulfide, a fair semiconductor, capable of mak¬
ing a rectifier, /rap)

Anyway, at that time I lived in Hackensack, N.J., just off Route
17, about a mile from where a 50-kW AM station on 770 kHz
resided. The station was then called WJZ, the flagship station of
ABC. During the summer, when it was hot, I used to leave the
door open, with a screen door keeping out the mosquitoes that
N.J. is noted for! The first night when I went to bed. I thought I
heard some music and people talking. Using my ears as a direc¬
tional antenna, I moved around until I found the source. The

screen, in the screen door, was vibrating and causing, I guess,
rectification of the RF comirg out of WJZ! Wearing a pair of
headphones, I poked around with the pin-jacks on the head¬
phones, and I found a place on the mattress spring where I
could connect the jacks with clip leads. I was able to fall asleep
every night to music without a radio and batteries!

Diran Varzhabedian (via e mail)

• Pease : Thanks for the e>cellent story. I've heard o* people
whose tooth fillings can similarly rectify out radio sigrals.
and they can hear the radio in their head. This could he
used to fake psychic seances. OO

(If you wondered about the headline. Happy April Fools!/rap)

Comments invited! rap <igalaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

^electronic design, Feb. 17. o. 20

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., San'a Cara, Calif.

• 18 •

03.31.05 Electronic Design

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHAT’S ALL THIS “WOMAN
SCIENTIST” STUFF, ANYHOW?

After the president of Harvard, Dr. Lawrence Summers, got in
hot water for questioning why so many women may not have
the aptitudes for science, now it is my turn to put in my two-
cents worth. The aptitude of women to be good at science, or
engineering, is well documented to be at a lower rate than it

is for men. This has been widely pub¬
lished by the Johnson O’Connor
Research Foundation (www.jocrf.org),
which has been studying aptitudes
since 1922.

If you take the top 1/4 of men in this
aptitude, many of them are the
extremely good scientists and engi¬
neers. Then take the top 1/16 of
women-they are just as sharp at these
physical relationships, which the JOCRF
calls "Structural Visualization" (S/V),
with a similar distribution of excellent
skills. I mentioned this in my column on
“Aptitude Stuff" (electronic design, Nov.

3, 1997, p. 219). (Also see www.jets.
org/latestnews/JOCRF-article.cfm.)

Now, not every kind of science or engi¬
neering requires this S/V, but many do.

So I tend to support that poor belea¬
guered man, Dr. Summers. Hejust
happened to have the bad luck to be
the president of Flarvard while he was
telling the truth, or at least asking the
right question. Give or take a few percent, 1 fnln’K 'ne is largely
right. Some people then bleated, “But there are many women
who excellently head science departments at many major
universities." One cannot argue against these facts, as they
are facts.

But just because a woman is running a science department, it
does not necessarily mean she is a great scientist—and it does
not mean she is not a great scientist. Rather, it means that she is
a good administrator and is good at getting along with the wild-
men (and women) scientists. That by itself is an admirable art,
and a very good talent, but not necessarily directly related to
being good at science. This is sometimes called "herding cats."

It's too darn bad we all have to be so (bloody) politically cor¬
rect these days. And, it’s too darn bad that some people have
to be so touchy about situations that have been documented
as factual.

Also, note that some studies show
that men who are good at engineering
and science get those traits from their
mothers. It is not yet well stated where
the great women scientists get their
aptitudes from.

Was my mother very scientific? I don’t
ever recall debating her on science. But
my mother, Beulah K. Pease, was a
schoolteacher for over 40 years, and I
would say that her common-sense atti¬
tudes were never far from science. And I
score at the 99th percentile-or higher-
on all the S/V aptitude tests. My wife and
both sons are very high at this S/V, also.

Meanwhile, all of us who are good at
science and engineeringshould go out
of our way to help any kids who are
good at science and/or math—or,
indeed, whatever they are good at. We
really must encourage young people,
and especially women and minorities,
when they show a spark of brilliance
and natural aptitude.

Girls are often d'iscouraged from taking careers in math or
science, and we should not let that happen. I can’t easily nor
seriously encourage (by example) a kid at drawing or music,
but I can do something for kids in science or math. Give her (or
him) a Heathkit!! (More on this later...)

Do you folks all agree? Or do you think I should apologize
(like Dr. Summers) for what I said here? Fat chance! GO

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

Read the Rebuttal: Editor-in-Chief Mark David has invited the IEEE's Women in
Engineering to respond to Bob Pease's viewpoints (which do not
necessarily reflect the opinion of Electronic Design or Renton
Media). For the rebuttal, go to elecdesign.com, Drill Deeper 10117.

• 20 •

04.14.05 Electronic Design

• •

PeasePORRIDGE

Jill S. Tietjen

Yes, Engineering Is A Woman's Job

As a woman engineer, I have worked for over 25 years to encourage more girls to pursue careers in science, technology,
engineering, and mathematics (STEM). I had to stifle my screams and overcome my first inclination to label both Larry
Summers and Bob Pease as Neanderthals (or worse) after reading Mr. Pease's column in the April 14 issue of Electronic
Design. Instead, I decided to apply my analytical reasoning skills (a highly desirable aptitude for pursuing an
engineering career) to Mr. Pease's arguments to see if there was any "there" there.

I found that intrinsic aptitude differences between the sexes are not that significant. The scholarly research on the whole
question of nature versus nurture has found that gender differences on science and math tests are small and decreasing
and that a significant number of complex and not well understood factors, including expectations, influence
performance. 1 If genetic and intrinsic aptitude explained why women still constitute such a small percentage of the
STEM workforce, how does one explain the increase in female, undergraduate students studying engineering in the
U.S.? In 1970, 0.3% of the degrees awarded nationwide went to women. Now women earn almost 20% of the
engineering degrees. This is an increase in magnitude of over 60 times. Surely women haven't improved their
quantitative skills by this magnitude just in the last 30 years. Factors other than aptitude must account for this result. 2

The National Academy of Engineering's (NAE) treatise, The Engineer 0/2020: Visions of Engineering in the New
Century, commented on the rapid pace of technological change. The engineering profession will be driven in the next 15
years, this report predicts, by breakthrough technologies including biotechnology, logistics, high-performance
computing, materials science and photonics, and nanotechnology. To function in this quickly evolving world, future
engineers must not only possess critical thinking skills but also strong analytical skills, practical ingenuity, and
creativity and innovation. The engineer of the future also has to be dynamic, agile, resilient, and flexible; communicate
well to diverse, global audiences; be able to master business and management principles; understand and practice
leadership principles; hold high ethical standards; possess a strong sense of professionalism; and be a lifelong learner.
What are the skill levels of the average male for these characteristics?

These skills somewhat parallel the aptitudes that Mr. Pease alludes to in his article — structural visualization, analytical
reasoning, memory for design, and mathematical ability. But much like veterinary medicine, engineering is changing as
technology and the information age change our world. Veterinarians no longer need to be big, strong men who handle
farm animals. And engineers no longer require a primary aptitude of structural visualization, as did the white males
who constituted almost the entire engineering workforce until the mid-1980s and still constitute the vast majority of
engineers. In fact, structural visualization is not a characteristic that the NAE identifies as an important skill for the
engineer of 2020. Most engineering disciplines no longer even require a drafting class like the one I took in the mid-
1970s when I was in college. The advent of the PC and computer programs that produce three-dimensional images
changed that.

Women excel in many skills that the NAE projects the engineers of 2020 will require. Women are multitaskers,
networkers, communicators, project managers, and problem solvers. Women see things differently than men. Women
would not have designed the early versions of VCRs that no one could program. Women would not design a PDA
specifically or primarily to fit in a man's breast pocket. Plus, women would not design car airbags that are likely to
seriously injure or kill people when it deploys. But 30 women were involved in the design of the Ford Windstar. This
very popular van won the five-star governmental award (the highest) for frontal crash tests. In addition, it included such
features as square cup holders designed for juice boxes and a reverse sensing system that beeped to warn backing-up
drivers of objects in the way. It also had sliding doors on both sides that could be opened by clicks on the key fob. 3
Women think differently.

It doesn't make sense to me to purposely or unintentionally discourage half of the population from pursuing a STEM
career, especially when this type of career provides such value for the world's population and can be so satisfying. We
are discouraging the female half of the population. One of my coworker's daughters was having trouble with first grade
math in a major metropolitan-area public school. This girl's teacher and principal both told her parents, "Don't worry. It
isn't important for little girls to do math." Many math- and science-phobic teachers and mothers think math and science
are "hard" and thus teach their female students and daughters to avoid math and science. There are also those parents
who choose only to pay for higher education for their sons and leave their daughters to pay for and obtain any higher
education themselves. The discouragement results in girls who play dumb to be popular and then don't enroll in the
advanced math and science tracks in eighth grade that will lead them to STEM careers.

April 14, 2005 Electronic Design

PeasePORRIDGE

Yes, Engineering Is A Woman's Job

Jill S. Tietjen

to let the large insults that result in a hostile environment slide (being choked on the job site by a client, having a
coworker throw a refractory brick at you).

If Larry Summers truly wanted to provoke discussion on the topic of the paucity of women in the STEM workforce, he
has succeeded beyond his wildest dreams. The bottom line, however, is that the global economy needs women in STEM
careers, and it needs to tap their innovation and creativity and their different way of thinking. Certainly, if women can
handle medical, accounting, and legal careers, then science, technology, engineering, and math are also jobs for women!

1. "Response to Lawrence Summers' Remarks on Women in Science," Women in Science & Engineering Leadership
Institute, University of Wisconsin, Madison.

http://wiseli.engr.wisc.edu/news/Lawrence_Summers_Response.pdf

2. "Women in Science: What are the obstacles?"
http://whyfiles.org/220women_sci/index.php?g=2.txt

3. "Windstar Moms Team Engineers Parent-Friendly Design"
http://www.autoworld.com/news/Ford/ParentFriendly.htm

April 14, 2005 Electronic Design

• •

PeasePORRIDGE

Jill S. Tietjen

Yes, Engineering Is A Woman's Job

The discouragement continues in the STEM workplace. Here, women who stay in STEM careers do so by battling past
or learning to ignore the small insults that occur regularly. (You're pretty good for a woman engineer, you look like a
woman engineer, you're pretty smart for a girl, what's a nice girl like you doing in a place like this?) Women even learn
to let the large insults that result in a hostile environment slide (being choked on the job site by a client, having a
coworker throw a refractory brick at you).

1. "Response to Lawrence Summers' Remarks on Women in Science," Women in Science & Engineering Leadership
Institute, University of Wisconsin, Madison.

http://wiseli.engr.wisc.edu/news/Lawrence_Summers_Response.pdf

2. "Women in Science: What are the obstacles?"
http://whyfiles.org/220women_sci/index.php?g=2.txt

3. "Windstar Moms Team Engineers Parent-Friendly Design"
http://www.autoworld.com/news/Ford/ParentFriendfy.htm

April 14, 2005 Electronic Design

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB’S MAILBOX

Hi Bob: I see more and more uses of LEDs in place of incandes¬
cent lamps every day. However, many of our engineering
brethren are not looking at the big picture with these. It would
appear that they are not considering the devastating interfer¬
ence that some of these replacement LED (power-supply)

units are creating on the AM radio broadcast bands.

For example: traffic signals. While I was waiting under a red
traffic signal, I had a barrage of repetitive static wiping out the
AM broadcast station that I was listening to. That is, until the
light turned green. This RF buzzing interference had not hap¬
pened before at this location. Upon closer examination the
next day, it appears that (just) the red traffic signal incandes¬
cent bulbs were replaced with a load of red LED signal lamps,
hence the interference during red signals, and not during the
yellow and green incandescent illuminations. (Apparently,
some guys have used crude switchers to convert high V to high
current, and to hell with the RFI/EMI. Quite unfortunate, /rap)
In summary, more thought should be put into our design
projects to ensure that we don’t cause problems with other
existing technologies, and in this case, we need to be aware of
working on LED replacement lamp power-supply solutions that
don’t cause major interference. (For sure!!/rap)

I doubt if these red LED signals would meet FCC Part 15 reg¬
ulations, as I would have definitely missed a tornado warning
broadcast on that station until I was able to drive away from
the offending RF generator. (I'll switch over to AM reception.
Here in the Sunnyvale area, there are a lot of LEDs, /rap)

The other item down the pike is the proposed Broadband
over Power Line (BPL), which has the potential to derail com¬
munications across the board. But that’s another topic for
another day.

• John Pavlica (via e-mail)

• Pease: That’s a whole other mess. All my ham friends are
furious about this proposal.

Hi Bob: Just read Jason Cook’s inquiry on eyelet usage in pc
boards (electronic design, Feb. 17, p. 20). It sent me back to the
early ’70s when I serviced consumer Hi-Fi equipment. A well-
known manufacturer (KLH) produced the first reel-to-reel tape
recorder with an integrated Dolby system. Technically, the
machine was quite an evolution and sounded excellent. Howev¬
er, the vast majority developed intermittent problems virtually
immediately. The KLH Model 41 was plagued by the use of
“eyelets” through the pc board to connect upper and lower lay¬
ers. I’ve spent many an hour chasing down intermittent con¬
nections in KLH Model 41s. All this to say—eyelets? Please no.

A piated-through connection is the better way to go. Surely,
there was nothing saved by the use of eyelets if reliability was

any concern at all. (It is imaginable that a really heavy compo¬
nent might make you wish you had an eyelet to support the
heavy weight—in shock conditions—which would, of course
cause them to go open!/rap)

• Geoff Pomeroy (via e-mail)

• Pease: Hi, Geoff, your story is consistent with other people's
stories. If I had a board without eyelets, I would just wire a
wire through the hole and solder the wire (at least 3/16 in.
of it) to the top and bottom foils. Or I would preferably do
two of those, in parallel, through adjacent holes. I’ve done
that. I am convinced this is very reliable. If I had a pc board
that already had an eyelet, I would solder a wire through the
hole (or solder it to the component lead that goes through
the hole) and extend it about 1/4 in. away and solder it, top
and bottom, to the two foils. The 1/4 in. would act like a
strain relief. Is this an adequate fix or “band aid”? it’s con¬
sistent with what I have seen, but I repaired a lot of my pc
boards, and almost none of anybody else’s, with eyelets.
Philbrick used eyelets, in Matter’s SP656, in parallel with
shorted (top-bottom) edge connectors, so we would never
see any failure. Failures would be unnoticeable. ©©

Comments invited! rap@galaxy.nsc.com —or:
Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara. Calif.

• • 20

04,28.05 Electronic Design

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

* i

WHATS ALL THIS NAP
STUFF, ANYHOW?

I like to sleep. I like to wake up. But I don’t like to wake up at 1:40
a.m. and not be able to get back to sleep for a few hours. Two
o’clock, 3, 4... and this lying in bed is not restful. Finally I may
get back to sleep at 5 and when I wake up at 7, the newspaper is
waiting for me to read it. I’m not off to a good start on the day.

I think I’ve solved this
problem. When I wake
up at 2 a.m., I wait for
barely 1/2 hour, and if I
can’t get to sleep, I get
up. I read some e-mails
or other mail, and I
type, and perhaps write
a column (as I am doing now). By 5
a.m., I can usually get the newspaper.

If I then get a “nap attack.” and go
back to sleep, that is great.

I can wake up by 7:30 a.m., surprisingly well
rested. Trying to get back to sleep at 2 a.m. can
be a loser. Big waste of time. Getting up and
working, to get sleepy, is not a terrible idea. I
don’t know why I can’t get back to sleep at 2 or 3
a.m. I guess I am rested just enough that I can’t
get back to sleep.

I never have much trouble with jetlag. When 1 fly
at 4 p.m. to Europe, I may fall asleep on the plane, but not get much
rest. So by the time I get to London-Heathrow (LHR), I am a little
tired. However, by 5 p.m., I get quite sleepy. If I fall asleep then, I
might sleep for several hours and then wind up wide awake in the
middle of the night! That is the bugaboo of jetlag-not being able to
get your sleep/awake cycle into the right phase. So:

(A) If you get sleepy at 5 p.m. on your first day in Europe-go
ahead, goto sleep—but set two alarms, as needed, to wake you up
after just a couple hours. Then, get up and do the town, and eat and
drink, all evening.

(B) Drink is the other part of the solution. Drink lots of beer, or
whatever, and stay awake and cheerful until at least 11 p.m., prefer¬
ably later. Then go to bed. You are then so tired, and well hydrated,
and full of alcohol and food, that you can now sleep to morning. This
is like taking a phase-locked loop and forcing it to a new phase.

Sometimes I don’t get enough sleep during the week. I build up a
sleep deficit. When I get to Saturday, I sleep late, and I am still tired.
Obviously I need to get another nap.

Once, I was driving down toward work. I was quite sleepy, but I
kept awake just fine all the way down to El Camino Real. Then I

had to make a left turn onto Shore¬
line. I stopped in one of the two
left-turn lanes. A few
minutes later, a woman
came by and rapped
on my window. “Are you
all right?” I had fallen
asleep while waiting for
the green arrow. I figured,

I might have to take a nap
later in the day—but I did¬
n’t. Apparently the few min¬
utes of nap (sitting parked in
the middle of the road) kept
me rested enough for the rest
of the day. I was a bit surprised
by that.

j As Isay in my pretty
good book*, driving while
sleepy is pretty dangerous
and com parable in danger to driving under the influence of alcohol.
Washing your face with cold water, or taking a snack with caffeine,
or singing loudly, may work for a little while. But when these do not
work, take a nap. It’s dangerous not to.

We’ve all heard the warning: “If it is very cold, don’t go to
sleep, or you may never wake up.” But this is not necessarily
true. Napping when you are only moderately sleepy is probably a
good idea. Staying awake until you are exhausted is probably not
a good idea, as you may be more prone to chilling when you fall
asleep at that stage. OO

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090 , Santa Clara , CA 95052-8090

*How To Drive Into ACCIDENTS—And How NOT To, page 277,
Pease Publishing, at www.transtronix.com.

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

• • 20 • •

05.12.05 Electronic Design

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB’S MAILBOX

Dear Bob: I have an odd problem. Early British cars used an
electromechanical vibrating regulator to supply -10 V to the gas
gauge from the 12- to 14-V battery (“positive ground”). The
current demand is low, probably 1 A would do. I have been, so
far, unsuccessfully searching for a -10-V regulator in a TO220

package. A positive regulator would be trivial to find. I did find
an MSK part, but they wanted $100. New mechanical devices
are so poor that many are DOA and others have a short life.
They cost about $25.

« Gene Mallory (via e-mail)

• Pease: Hello, Mr. Mallory. Look up the LM337 and see if it
won't do your job. It will need 124 El, 1%, and 866 El, 1% (or
100 El and 700 El, 1%). And there you have a -10-V output.
It will also want 10 or 20 jiF of output bypass capacitance
for loop stability and to suppress transients. And if the input
source is faraway, it will need a couple of microfarads at
the -14-V input. And the case of the LM337T will be at -14
V. So, you might need an insulator to couple it to your
heatsink. The going price in Akihabara, in Tokyo, is about 79
cents, or 89 yen. They are also stocked by Digikey or similar
distributors. (You DID ask the right guy. I designed the
LM337T about 28 years ago—and it’s still in production.
SOLID, /rap)

Dear Bob: While reading your conditioning on the stairs arti¬
cle (electronic design, March 3, p. 20), I was reminded of a
friend whose mom conditioned him and his siblings indoors.
She put backpacks filled with rocks on her children, then
marched them through the house day after day. When she
thought they were ready, she marched them out of Lithuania to
freedom back in the forties. I think I must have just sat there
with my jaw dropped while Ivar told me the story of how he
walked to America. Gives homework a whole new meaning.

« Peter Nord (via e-mail)

• Pease: Hello, Peter. Wow, I am impressed! I read the story of
a man and his son who walked from eastern Poland across
Europe in the 1930s. When they got to France, the man
died, but the son kept going to get on the boat to the U.S.
Same idea and also very impressive!

Dear Bob: I never really thought about using the one flight of
stairs in my house for conditioning—although I generally climb
the steps at work two at a time with only a light grasp of the
handrail, as I don’t want the safety geeks to get upset. (Here at
NSC, some "safety geeks” put up a sign to "always use the
handrails to avoid accidents ." After I see that sign, I never use
the handrails! I always run up and down stairs. But I rarely run
down two steps at a time, /rap)

I like bicycling and am convinced this will help build my legs for
the hills around here. However, my comment is more aimed at

the maximum heartrate issue. I always heard 220 minus your
age is an absolute maximum. (You might be quite right, but 200
-xisa good recommended maximum rate, /rap)

Several years ago, after taking a spinning class at a gym and
seeing my heartrate hit 175 bprn (at the age of 50), I asked the
instructor about this max heartrate thing. She replied that 220
minus age was a good rule of thumb, but people who have reg¬
ularly exercised for a number of years can usually attain higher
readings than the formula predicts.

In case you think I’m just patting myself on the back, I also
believe that like any rule of thumb, individual differences exist
regardless of other factors. Thanks for the tip on the stairs. I’ll
probably incorporate it in my own workouts on days off the bike.
• Barry Cantor (via e-mail)

» Pease: Wow. I’ve done 160 bpm after a long uphill pull, hiking
at the age of 60, and I thought I was pushing it. When I walked
away from all my buddies back in 2000, at 18,000 ft, I did not
have any measure of how fast my heart was beating, but I was
definitely limited by how much air I could bring in through my
NOSE. I was probably near 160. OO

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, Calif.

• • 20 • •

05.26.05 Electronic Design

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHATS ALL THIS DOCTORING
STUFF, ANYHOW? (PART 4)

Usually, I write my own columns. But recently I received a letter
that was just too good to include in a “Bob’s Mailbox.” So I
decided to make it the latest installment in my series on self-
doctoring. Thank you, Mr. Kraus, for taking the time to share
your hard-earned insights with our readers, /rap

Dear Bob: I had carpal tunnel, bipolar, migraines, body aches,
leg cramps, and crippling prostatitis. My doctor tried to prescribe
(respectively) surgery, drugs, nothing, nothing, nothing, and lots of
sex. Given the progressive patterns of my symptoms—as I experi¬
enced them—I suspected a food-related systemic degenerative
problem. And I wanted guidance from my doctor. My doctor would
hear nothing of it, because “food intolerances are rare.” But he
warned me about a teensy elevation in my cholesterol.

Well, inspired by your self-doctoring articles (electronic design,
Sept. 30, 2002, p. 110; Oct. 14, 2002, p. 86; April 12, 2004, p.
18), I studied food intolerances on my own. And to my joy, I have
successfully treated these problems and more by eliminating any
source of gluten. I used to be derisive of diets like this. Now I’m on
the other end of the derision—karmic, eh?

There is a lot of public doubt on the subject. I wouldn’t have
believed it myself if I lacked the scientific background to read the
research or keep an engineeringjournal of my symptoms as I
tried the diet. My doctor is still cool to the idea, so I finally
pestered an out-of-state friend (an MD) for support. He recently
sent me this note:

“No one but one of us who has suffered from gluten intoler¬
ance can really understand what you have been through for diag¬
nosis, or the victory you have won . Don't forget, you already diag¬
nosed this yourself, and just asked me for information to confirm
your own hypothesis. You cured yourself!

"It brings me huge joy to hear that you have healed and found
more joy in your life, and are not a victim of your genetics and
well-meaning but not-well-informed doctors.

"/ saw a patient who I and all her doctors had thought had MS
for 15 years, who was dying by inches, have a dramatic turn¬
around last month on a gluten-free diet. I wept that I had missed
her food intolerance. You owe me nothing. You are a friend. Just
be alert, and someday you can pass it forward ."

Self-doctoring seems to be the rule with gluten intolerance. U.S.
doctors have been taught that it is only a rare childhood disease.
But 21 st century research proves that gluten intolerance is very
common and has many symptoms that are easily mistaken for
other conditions. Gluten’s strong contributions to other conditions
such as diabetes and depression are only recently under study.
Conventional blood tests have been recently shown by Kenneth

Fine, MD, (www.enterolab.com) and others to produce more false
negatives than true positives. Thus, many seriously sick patients
have been told that tests show that nothing is wrong with them.

Doctors have hundreds of patients and only get the odd 15
minutes here and there to talk with you, but you are with yourself
all of the time. You ought to quarterback your own health care.
Lists of gluten intolerance symptoms, tips for diagnosis, and links
to research papers and Celiac sites are available at www.mem-
bers.cox.nel/barold.kraus/gluten.htm. With sincere gratitude,
Bob, for your self-doctoring essays,

• Harold Kraus Jr. (via e-mail)

• Pease; The occurrence of Celiac disease used to be estimated

at a few ppm. Now it’s recognized as around 1%. If you have
symptoms, and no cause, consider a three-week vacation from
gluten. Thanks, Harold, /rap ©© ED Online 10410

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at Nation¬
al Semiconductor Corp., Santa Clara, Calif.

.. 20 •*

06.09.05 Electronic Design

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB’S MAILBOX

copper lead frames unless they are some kind of supercon¬
ducting alloy.) The copper wires in my house are 14 gauge and
are only rated for 15 A, and they are bigger. And what about the
die bond wires? Is someone pulling my leg or what? Is this a
new form of specsmanship?

• Ed Ganshlrt (via e-mail)

• Pease: Hello, Ed, #14 copper wire is rated to fuse at 160 A
(per the CRC Handbook). So there's a safety factor of 10, for
surges. But a TO-220AB’s lead is just 18 mils thick by 45 mils
wide, tapering to 28 mils (min), and the datasheets say it's
good for 75 A? I can't believe that! Let's try it! Stand back!

HI Bob: I am writing to express my concern over what I regard
as premature adoption of lead-free soldering in our industry.
There seems to be an almost lemming-like quality in the rush
to adopt this technology. (Yeah, /rap)

I have yet to see any concrete data regarding manufactura¬
bility and reliability, although the consensus appears to be that
both are going to get worse. Would you want to rely on a defib¬
rillator that used lead-free solder? (No!/rap)

• Ken Neltnor (via e-mail)

• Pease: Hello, Ken. NSC's MIL AERO group is keeping the
lead-based solder, that is solder-dipped onto the leads of hi-
rel parts. Some examples are LM108AH-883 or MIL-38510
parts for military and space applications. Similar industrial
parts (LM108AH) will have solder-coated leads without the
element lead. It will not be a straight tin solder, as many oth¬
er 1C makers are doing, but we will have a tin-silver-copper
coating that is dipped on. And that is apparently better for
whisker-free manufacturing. But at any time, if you need the
highest reliability you can get—such as for pacemakers or
defibrillators (even with inexpensive plastic-packaged
parts)—you can use conventional solder with lead. That is, if
and when you can get it. That will prevent the tin-whisker
problem. Talk to your favorite solder maker, and get his opin¬
ion. Okay? Good question, and thanks for asking.

Hey Bob: Yesterday, I took a break from the breadboard of
my latest design to graze through the new copy of Electronic
Design. Dick Weiner’s letter was nice to see (electronic design,
June 23, p. 18.). I’m not the only person who wants to actually
see his designs work. Makes it a lot easier to defend yourself in
the meetings.

What I was taking a break from was installing eight SOT-23
transistors onto a 24-pin Aries SOIC board. They work just fine, if
you have the patience to solder them into place. (Actually, a cou¬
ple of these are gonna be used as diodes. Thanks to you for
that.) You can use the transistors, diodes, Zeners, and such that
you’re actually going to use in production. (True, but the adapter
board does add a little bit of stray capacitance and inductance.
Gotta watch out for that!/rap)

I design voltage regulators and similar circuits for my compa¬
ny’s line of heavy-duty automotive alternators, so I don’t have to
deal with speeds above the clock on a microcontroller. There¬
fore, speed is not an issue. These Aries SOIC boards can even
be used for five-pin SOT-23 parts. Takes a little imagination and
patience, but they work just fine for the little 7101s and 7301s.

• Edward Craig (via e-mail)

• Pease: Glad you're having fun with the tiny circuits, Edward.

OLD PHILBRICK AMPLIFIERS • My old friend Tom Decker
has some old Philbrick Amplifiers that he wants to be selling
soon, such as K2-Ws and USA-3s. You could look on eBay or
write to him at New Dimensions, 978 West County Road I, St.
Paul, Minn. 55126-1315. ©O ED Online 1090$

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, Calif.

08 . 18.05 Electronic Design

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB’S MAILBOX

Hi Bob: Does anyone breadboard circuits anymore? In the good
old days, just a few years ago, you used to draw circuit dia¬
grams in your articles that looked like something a college pro¬
fessor would put up on a white board. We design engineers
would all run out to the lab and try it out in a couple of hours.

I always seemed to get more out of actu¬
ally trying out my more complex circuits
than from simulating them.

However, today a lot of the newer, excit¬
ing ICs come only in such small packages
that even a neurosurgeon would be hard
pressed to hook them up. Have you ever
tried to solder wires to a quad op amp in a
14-pin, small-outline 1C (SOIC) package?

Have you seen the packages that these
three-pin precision voltage references
come in? If you sneeze—game's over? I’m
all for miniaturization, but not in the
breadboard stage. Not alt of us are work¬
ing in the gigahertz range where lead
length is critical. Have you ever soldered
#36 wire to one of these parts and then to
an old-fashioned DIP socket—so that you
could wire wrap to it?

I recently got a sample kit of op amps
from one of the major 1C vendors. It has
some really neat parts. The largest pack¬
age was an eight-pin SOIC. To add another level of complexity,
they included tape and reel parts. I mean, come on, how do
you build breadboard circuits with fly-spec parts? (W/th perse¬
verance—or with adapters, /rap) Are there any inexpensive
chip carriers that convert these parts to fit into an old-fash¬
ioned DIP socket? Does anyone out there have a solution, oth¬
er than laying out custom pc boards every time you want to
breadboard a circuit?

• Dick Weiner (via e-maii)

• Pease: Yes! Let me find you the standardized answer from
Paul Grohe. DigiKey carries them. For non-gigahertz applica¬
tions, these are good, and they have solder-mask, so they
connect up and don't short out.

• Paul Grohe: Hi, Dick. For SO-8/14-to-DIP conversion, <100
MHz, the Aries “Correct-A-Chip” series is what I use. The
adapters seem pricey at first, but because they are solder
masked, they are easier to solder and desolder and can be
reused several times (www.arieselec.com/products/
correct.htm). Models 08-350000-10 and 14-3500000-10
are available from DigiKey and Mouser : / have used these
well into the hundreds of megahertz with op amps (with the

expected peaking), but I would not try
them on fast, high-current switchers or
other LC RF-critical applications where
extra lead inductance and capacitance
could be lethal to the circuit But for low-
power, low-speed apps, they are all but
invisible.

For SOT-23, SC-70, MS OP-8, and other
nasty little “transistor” packages, the
Capital Advanced Technologies “Surf¬
boards ” can be used (www.capitalad-
vanced.com). These adapt SOT-xxx to a
0.1-in. spacing SIP. They do not have sol¬
der mask and don't tend to last as long,
but they are a bit cheaper. They are avail¬
able from DigiKey, Newark, Fry’s, and
Jameco. By the way, you can “stretch” an
SC-70-3 onto a SOT-23-3 footprint. So
the SOT-23 adapter can do " double-duty”
as an SC-70 adapter, too.

Hi Bob: I tried your stair-climbing idea
(electronic design, March 3, p. 20) this morning instead of my
usual Sunday morning outdoor downhill/uphill plod, as it was
raining a bit here near Seattle, Wash. (Imagine that!) 1 decided
to start with 50 laps and it took me about 20 minutes. Nice lit¬
tle workout. (Fifty is a pretty good start. When I was out of
shape, I did 65—which is lousy—but it was a start, /rap) The
only comment that I have is that I prefer a little 10-minute cool¬
down walk afterwards, just to gather my wits. Thanks for the
suggestion.

• Dave Gerstenberger (via e-mail)

• Pease: My preference is to just lie there, but I guess I am

used to that. While I am lying there, I can count my pulse
rate (up near 150 bpm?), then wait two minutes and see if
the rate dropped a lot ©© ED Online 10521

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara , Calif.

•. 18

06 . 23.05 Electronic Design

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHATS ALL THIS FLOOBYDUST
STUFF, ANYHOW? (PART 14)

This year I’m going to start off with magazines: It is well known
that huge (astronomical!) quantities of magazines —especially
National Geographic—have, been stored in a million attics
across the U.S. This raises the center of gravity of the house
and the Polar Moment of Inertia of the whole Earth.

Physicists have predicted that the Earth will wobble excessively,
its rotation will slow down to a critical speed, and then the Earth
will fall out of its orbit sometime in the first half of this century. It
is not known if this will happen earlier than the demise of the
Social Security Trust Fund, or later—or simultaneously?

SCIENCE FICTION • My friend Bill Bernardi recommended to
me a good science fiction book by Eric Harry: Society of the
Mind. Bill wrote, “Hey Bob, I just read... this book. It reminded
me often of you.... It’s about this young genius who can’t han¬
dle the real world, so he buys an island and stocks it with the
best and brightest people of the time, from all fields... pays
them millions of bucks.... His philosophy is that the digital
world of computer technology is a dead end and must evolve in
analog form to continue to grow. So he used neural networks
and fuzzy logic to let the analog computer teach itself how to
grow bigger and better and smarter....”

I read it, and I agree that this was a fascinating, fanciful, and
enjoyable story about how one brilliant man could plan one big
computer that could build itself bigger and better—until some
very interesting things happened. I’m going to buy a few sec¬
ond-hand copies of this 1997 book to give to friends.

SCIENCE HYPERFICTION • I had to cobble up a simple clock
and a 74C163N counter to make that stair-step waveform for the
“conditioning” column (electronic design, March 3, p. 20). So I
grabbed a solderless breadboard, and I lashed up the circuit—for
the first time ever. All the wires were flaky and confusing (not any
choice of color, for a given length of wire). The resistors were quite
wobbly and flaky. The insulating space between adjacent compo¬
nents was marginal. I got it to work Gust barely), but the connec¬
tions were not reliable. I do not like solderless breadboards. This
was a setup—an experiment kit for students—that Forrest Mims
sent me. I like his experiments, but I just don’t like using a solder¬
less breadboard to connect them. My favorite programming lan¬
guage really is solder. The experiment kit, Learning Lab Model 28-
280, costs about $60 at any Radio Shack (www.radioshack.com).

GOOD NEWS! • I told you a few years ago that across the
world, all makers of polystyrene had stopped making the materi¬

al, and capacitor makers were going to run out of polystyrene. I
have been able to use polypropylene for almost all of my preci¬
sion capacitor needs, with no degradation of accuracy. But now
in the U.K., there is a company selling polystyrene capacitors. If
you can tell the difference between polystyrene, polypropylene,
and any other kind of “poly” caps, you might want to buy some
polystyrenes from LCR Capacitors (EU) Ltd. Unit 18 (www.icrca-
pacitors.co.uk).

DOCTORING, PART 4-B • A friend said that his daughters once
got really sick with strep throat. After treatment with antibiotics,
they got better quickly and went back to school. But shortly, they
got sick again, and cured again, and sick again, and again. The
doctor tried to figure out how. He tested the mother-was she a
carrier? No. Neither were the father and teacher. The MD was
stumped—but not the triage nurse. She had their dog tested,
and the dog was the carrier. After the dog was treated, the prob¬
lems went away. (One clue was that the son did not get sick, and
the son did not play kissy-poo with the dog.) You can’t expect a
doctor to know everything. We don’t want to be unfair about
that. We have to help. ©O ED Online 10630

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara , CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Caiif.

• • - 20 • • •

07 . 07.05 Electronic Design

Pease PORRIDGE

( BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB'S MAILBOX

Dear Bob: I’m running into a head-scratcher. I am working on a
2 -jiA current source (pnp transistor plus sense resistor in the
emitter lead, and an op amp to control the pnp’s base so the
voltage across the sense resistor is constant) with as much
compliance to use as much of the 5 V as I have available.

(Constancy of current is important but I don’t care too much
whether it’s exactly 2 pA as long as it doesn’t have much temp-
co/drift.)

But I’ve only found one pnp transistor (PN5138) with any
beta specs for collector currents that go down that low. (The
PN5138 has—I saw-a guaranteed spec of 50 minimum at 100
pA. I did not see any guaranteed spec at 1 pA. This transistor is
from the Process 66, which is not nearly as high beta as
process 62. Good beta at low currents like IpAisa character-
istic-not a guaranteed spec . /rap)

Any idea why? I can see the manufacturers wanting to
test/characterize beta at the high end, but why are they leaving
off the low end? (It was very uncommon, even 20 or 30 years
ago, to see guaranteed specs for beta at 10,1, 0.1, 0.01, and
0.001 mA. Almost nobody wanted to pay for that. The 2N930s
and 2N2484S were npn types, and they were fully specified—at
high prices, /rap) Nobody seems to support discrete small-signal
bipolar devices anymore. (They never did want to. if you wanted to
buy 2N4250s, 2N2605s, or similar pnp devices in mil-spec for¬
mat, for $20 to $40, you could do it. But for your application, just
buy some pnp transistors with beta - 200 min at 1 mA. All mod¬
ern high-beta pnp types will have good beta holdrup at low cur¬
rents. Get something like the old 2N4250, PN4250, 2N2605, or
something similar. If you buy them at Fairchild, they would be any¬
thing with high beta such as 300 min at 1 mA, or 250 at 100 pA,
and they would be from Process 62. That was what they were
called when NSC made them, /rap)

I’d use a PFET instead of a pnp, but I’m told FETs begin to
leak at high temps with V GS = 0 (e.g. for NFET, 2N7000 specs
<1 uA at 25°C, V GS = 0, and <1 mA at 125°C, V GS = 0). And I
don’t have enough voltage available to apply a reverse bias to
V G s- Any thoughts?

• Jason Sachs (via e-mail)

• Pease: FETs are not guaranteed to not leak, but they cer¬
tainly are not guaranteed to leak. The manufacturers just
don't want to test below 1 pA of leakage, as that takes time,
which they hate to waste. Those Process 62s will make you
happy.

Dear Bob: I was looking at some back issues and I came
across your column on the selection of tubes in a Philbrick K2-X
(electronic design, Oct. 18, 2004, p. 16). It made me recall an
experience of mine. Way back in the middle of a long-gone centu¬

ry, I was designing a VHF television tuner. I was using a dual
tube, I believe a triode-pentode, specified by the manufacturer
for that purpose. We used a major brand, adjusted the circuit,
and found that it would work properly with any tube they sent us.

Then we tried another, equally reputable brand. No go. They
did not work properly. Both were specified for tuners. They had
the same published specs. But the specs defined only the oper¬
ation of each section. They did not specify the physical layout. It
turned out that the orientation of one section was the same for
both companies. But the orientation of the other section was
different. This caused a difference in the mutual inductance
between the elements of one and elements of the other. At 200
MHz, mutual inductance was crucial. (Uh, yeah!/rap)

This indicates the danger of relying on specifications that
cover only the major requirements of a product. These are ade¬
quate for most applications but omit “minor” details that may
be crucial for your particular application. On occasion l have
encountered the same problem with mil-spec integrated cir¬
cuits and generic prescription drugs.

• Matthew W. Slate (via e-mail)

• Pease: You’re exactly right, Matthew. There are many fac¬
tors that are not trivial. ©©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp ,, Santa Clara , Calif.

•. 20

07.21.05 Electronic Design

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHAT’S ALL THIS TYPING
STUFF, ANYHOW?

In the old days, it was so easy to try to type “3M” to get 3 M£2
in a Spice program, and of course you got 3 mQ. A typo error
can really waste a lot of time! Typing was a large part of our
job as a means of communicating to a computer. These days,
there are still many things to type, in addition to just typing

e-mails. I type these columns
and I write replies to people
who send me e-mails.

When I was a kid of about
10, my mother got a couple
of instruction books and
started to teach me to type
on her big upright Royal. It
was not easy for a kid to
learn, but not too bad. I nev¬
er did achieve the skill to
type a paragraph without
errors. I kept making an
error and going back and
restarting. In those days, we
didn’t have any "error correction.”

I was supposed to be doing touch-typing without looking at
the keys. Maybe I did almost learn that, but over the years, I’ve
noticed that I have to peek at the keys for much of my work. Of
course, if I just have to back up a couple strokes and then insert
an s, 1 don't have to look at the s. But when I was typing at my
computer in my dining room, I discovered that when it got dark, I
had to turn on the lights to type. I’m not a very good typist, nor
fast, but I can type about 20 or 30 words per minute after mak¬
ing corrections.

When I type, I usually keep my left fingers at the proper
home positions, “asdf.” But my right hand is more free to range
around, as I have to be prepared to hit the backspace or delete
keys or (with my right thumb) the left-click. My right middle fin¬
ger has to be busy, pushing around the tiny red joystick on my
laptop, in lieu of a mouse. So my ability to touch-type has been
further wrecked by the requirements for word processing. My
mother would be horrified! But it works.

Anyhow, I have learned to recognize quickly when I type a
mistake. With modern word processing, I find it easy to go back
a couple of spaces and make a correction almost instantly, and
without any need for thinking or reading what I have typed
wrong. That saves time.

I started out word processing on a Coleco Adam, and its
daisy-wheel printer was far superior to most cheap dot-matrix

printers of its day. The Fri¬
day I bought that Coleco
Adam, I was asked if I
could write a 2000-word
story about temperature
sensors. “When do you
want it?” Tuesday. “Yeah, I
think I can do it.” I had
enough basic facts, and I
cobbled together enough
text to mail in on Tuesday.

I graduated to PC Write
Lite (Quicksoft) on an oid
IBM PC. PC Write really did
have some nice features.
You could use PF8 to convert uppercase to lowercase, and vice
versa. This was very handy if you had inadvertently hit the caps
lock, and typed a lot of stuff as caps, by mistake. I miss that.
Later, I graduated to an IBM ThinkPad T42. I’m still not that
good a typist. On the other hand, I did wear smooth the right
hand of the space bar. I almost wore it out in three years. I typi¬
cally type about 1 Mbyte in a year-various papers, columns, e-
mails, and replies...

I’ve noticed that a good typist will type an with the ring
finger of the left hand. But most computer experts (and many
engineers) will hit the shift with their left little finger and reach
over with the right index finger to hit the No good typist
would do that. But I usually do. It doesn’t really slow me down
much. To this day, many engineers still use a hunt-and-peck
method. Do schools teach typing these days, or do engineers
just have to learn as they go? I’m not sure. It’s not the same
thing as writing software or playing computer games. ©©

Comments invited! rap@galaxy.nsc.com —or:
Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

08.04.05 Electronic Design

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB’S MAILBOX

Dear Bob: This is a general question about TO-220 packages. I am finding MOSFETs available in TO-220AB packages that claim to have
continuous ratings of 120 A. I can believe the die can take it if heatsinked well. But can those skinny 0.044-in. 2 wires take that kind of
current? (I think they are tin plated copper lead frames unless they are some kind of superconducting alloy.) The copper wires in my house
are 14 gauge and are only rated for 15 A, and they are bigger. And what about the die bond wires? Is someone pulling my leg or what? Is
this a new form of specsmanship?

Ed Ganshirt (via e-mail)

Pease: Hello, Ed, #14 copper wire is rated to fuse at 160 A (per the CRC Handbook). So there's a safety factor of 10, for surges. But a
TO-22oAB's lead is just 18 mils thick by 45 mils wide, tapering to 28 mils (min), and the datasheets say it's good for 75 A? I can't believe
that! Let's try it! Stand back!

Hi Bob: I am writing to express my concern over what I regard as premature adoption of lead-free soldering in our industry. There seems
to be an almost lemming-like quality in the rush to adopt this technology. (Yeah, /rap)

I have yet to see any concrete data regarding manufacturability and reliability, although the consensus appears to be that both are going to
get worse. Would you want to rely on a defibrillator that used lead-free solder? (No! /rap)

Ken Neltnor (via e-mail)

Pease: Hello , Ken. NSC's MIL AERO group is keeping the lead-based solder , that is solder-dipped onto the leads of hi-rel parts. Some
examples are LM108AH-883 or MIL-38510 parts for military and space applications. Similar industrial parts (LM108AH) will have
solder-coated leads without the element lead. It will not be a straight tin solder , as many other IC makers are doing , but we will have a
tin-silver-copper coating that is dipped on. And that is apparently better for whisker-free manufacturing. But at any time , if you need the
highest reliability you can get—such as for pacemakers or defibrillators (even with inexpensive plastic-packaged parts)—you can use
conventional solder with lead. That is , if and when you can get it. That will prevent the tin-whisker problem. Talk to your favorite solder
maker , and get his opinion. Okay? Good question , and thanks for asking.

Hey Bob: Yesterday, I took a break from the breadboard of my latest design to graze through the new copy of Electronic Design. Dick
Weiner's letter was nice to see (electronic design, June 23, p. 18.). I'm not the only person who wants to actually see his designs work.
Makes it a lot easier to defend yourself in the meetings.

What I was taking a break from was installing eight SOT-23 transistors onto a 24-pin Aries SOIC board. They work just fine, if you have the
patience to solder them into place. (Actually, a couple of these are gonna be used as diodes. Thanks to you for that.) You can use the
transistors, diodes, Zeners, and such that you're actually going to use in production. (True, but the adapter board does add a little bit of
stray capacitance and inductance. Gotta watch out for that! /rap)

I design voltage regulators and similar circuits for my company's line of heavy-duty automotive alternators, so I don't have to deal with
speeds above the clock on a microcontroller. Therefore, speed is not an issue. These Aries SOIC boards can even be used for five-pin
SOT-23 parts. Takes a little imagination and patience, but they work just fine for the little 7101s and 7301s.

Edward Craig (via e-mail)

Pease: Glad you're having fun with the tiny circuits , Edward.

OLD PHILBRICK AMPLIFIERS

My old friend Tom Decker has some old Philbrick Amplifiers that he wants to be selling soon, such as K2-Ws and USA-3S. You could look
on eBay or write to him at New Dimensions, 978 West County Road I, St. Paul, Minn. 55126-1315.

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090 , Santa Clara , CA 95052-8090

August 18, 2005 Electronic Design

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHATS ALL THIS TOASTER
STUFF, ANYHOW?

Well, I never owned a toaster, but I have owned several kinds of
toaster ovens. A long time ago, I had a good GE one. But then
Black & Decker took over GE, and while the toaster ovens
performed okay, they showed me poor reliability after a while.
They would fail shortly after the warranty ran out, and I’d buy

another. I soon got tired of this. They would either quit open-cir¬
cuit, or the heating element would go gablooey. (That’s a tech¬
nical term.) Either way, they were unrepairable. My wife volun¬
teered to buy the next one.

She went out and bought a new kind of toaster oven from
Black & Decker. When she got it home, it didn’t work at all. She
decided to try a DeLonghi toaster oven. It looks handsome, anc
it seemed to have many nice features. It's a model AS960, now
about two years old, and it does seem to be reliable. The only
problem is that it doesn’t toast

Whether you put in two pieces of bread, or six, it takes forever
to get them even light brown. I have to put it through two “Dark”
cycles to get it even light brown. I measured the power, using my
watt-meter (electronic design, May 13, 2002, p. 86J, and it was
barely 1200 W on our 112-V line. Not very healthy.

I decided to add a booster transformer to get it up to over
1400 W. I went down to Halted Specialties Corp., which had
many small transformers—and a few BIG HEALTHY power
transformers. Surely one of them would put out 10 extra V! I
got a few and doped out which windings would do what. I
wired three 115-V primaries in parallel and three 3-V second¬
aries in series. Lovely (surplus) transformers—this rig would
put out 14 A and stay cool.

I set up some standard bagel pieces. I set up my wattmeters
and voltmeters all around the toaster, along with the transform¬
ers. Sure enough, adding the boost transformers kicked up the
power from 1260 to 1470 W. The heating elements got red. I
prewarmed the toaster for three minutes, let it cool for one
minute, and toasted some bagel pieces at 1460 W for exactly
three-and-a-half minutes.

I let the toaster cool off for one minute and toasted a matched
set of bagel pieces at 1260 W—to get them exactly the same col¬
or. And the time was—exactly three-and-a-half minutes. H’mmm.
I’m not sure what went “wrong,” but I’m glad I ran the experiment
very carefully. Maybe this weekend I’ll buy a loaf of white bread
and repeat the experiment—1260 W, then 1460,1260,1460
W... I may learn a trick.

So if you buy a piece of consumer electrical equipment, don’t
be surprised at anything. Don’t even be surprised if you have to
add a little helper to Ohm’s Law.

The bagels toasted at 1460 W for 3.5
minutes are on the left; at 1260 W for
3.5 minutes on the right.

RELATED STORY • There was a guy, George Gobel, who wanted
to decrease the time for starting his charcoal broiler. He tried sev¬
eral kinds of boosters, chemicals, etc. Finally he hit on the ulti¬
mate charcoal starter: LOX. He would apply a small flame under
the charcoal. Then from the end of a 10-ft pole, he would pour a
few gallons of liquid oxygen over the charcoal. His camcorder
could not record any actual details, as the flame was too bright.
But in three seconds, the charcoal was hot enough to broil meat.
Think of all the time you have wasted trying to get charcoal start¬
ed. Now you will never waste more than three seconds! There is
one minor drawback: If you use a cheap metal grill, it will burn
right through. You may find that a porcelain or ceramic grill lasts
longer... Hey, maybe I should do that with my toaster! The URL is
www.doeblitz.net/ghg/.

Comments invited! rap@gaiaxy.nsc.com
—or add your comments online at ED Online 10940
—or: Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

• • 20 • •

09 . 01.05 Electronic Design

WHAT'S ALL THIS
STUFF, ANYHOW?

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WIRELESS

As Jaye P. Morgan sang 50 years ago, “They all laughed at Wilbur
and his brother, when they said that man could fly. They told
Marconi, wireless was a phony, it’s the same old cry.”"’ For over
60 years, I have been enjoying many wireless things. But these
days, people think “wireless” is a great new deal.

If somebody invents a wonderful new system with electro¬
magnetic radiation and digital coding, it must be good. “Radio’
is obsolete. “Wireless” is wonderful.

Just this morning, I heard Disney head Michael Eisner say that
before 1948, there were only three major movie studios, and no
television, for entertainment. He seemed to have
no inkling that there were several large (and lively)
radio networks broadcasting music, news, drama,
and entertainment all around this country (and
around the world) many years before World War II.

People don’t seem to pay no respect to radio and
other early wireless inventions. (Personally, I think
the pictures are better on radio.)

As you may have noticed, National Semiconductor is not a
major provider of digital VLSI chips that go into cell phones. So
you might say that NSC isn’t a player in this wireless market, eh?
No, you shouldn’t say that!

When the VLSI chips began to get specialized and optimized
for high performance in cellular phones, the clock frequencies
began to go up, and the logic feature sizes began to go down, as
did the power-supply voltages—just like for microprocessors. And
what falls out? A magic word! D/s integration.

When you make smaller features and run your VLSI chip on lower
voltages, certain functions are hard to do. A decent audio amp can’t
run efficiently on 1.8 V. Furthermore, a cell-phone amplifier has to
reject certain kinds of nasty noise. So a good audio amp for a cell
phone works a lot better on a separate chip at 3.3 V. Thus, this dis¬
integration makes an external amp a good idea.

We have many good amps optimized for various cell-phone,
portable, and wireless activities, such as the LM4880. (Why
do people want stereo amplifiers in a cell phone? Why not use
two cell phones, one for each ear? Don’t ask.) When your GSM
phone has a lot of “bumblebee noise” at 217 Hz, an amplifier
like the LM4890 can reject it nicely.

Switching regulators that can convert a wide range of battery
voltages efficiently to the needed regulated voltages are never
included on the VLSI chip. So we make lots of those regulators,
such as the LM3200. Others are used to pump voltages up to
go into LEDs and backlights for various displays. If you needed a

precision reference, you could integrate it on the main chip. But
the yield loss makes this a loser. Do the math. It’s cheaper to
add a separate chip. There are log-amplitude detectors for the
RF power output like the LMV225. There are power-amplifier
controllers, too.

Temp sensors for the voltage-controlled crystal
oscillators and the output RF power transistors—
and for battery recharging—usually do not go on the
main chip. The LM20 draws very low power, and
there are many other temp sensors optimized for
portable use, with analog or digital outputs. After
all, the main chip ain’t at the same temperature as
the crystal, which ain’t at the
same temp as the battery, a sin¬
gle temp sensor isn’t right to do
all those jobs. There’s a lot of
places where cell-phone
makers like to use linear
chips, just like the digital
computer makers do. They
just don't talk about it a lot.

So, is wireless really
“everywhere?” Not at
my house. GO

Comments invited!

rap@galaxy.nsc.com —or:

Mail Stop D2597A,

National Semiconductor
P.O. Box 58090, Santa Ciara, CA
95052-8090

BOB PEASE obtained a BSEE from MIT in
1961 and is staff scientist at National
Semiconductor Corp., Santa Clara, Calif.

* George Gershwin, “They All
Laughed,” 1937

RAP with his
latest, newest
wireless
telephone...

• • 22

09.15.05 Electronic Design

PEASE PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB’S MAILBOX

Hi Bob: I think this may interest you. I repair electronics and had a
guy bring me a TV/VCR. A month later he came back, and the
belt I replaced was broken. It was one of those take-everything-
apart-to-get-to-the-belt kinds. This time, the belt looked like it
was 20 years old—nothing but little, hard chunks of rubber.

junk. I feel sorry for anybody that has one in their garage.

• Pease: Oh, man!!! Think of all the rubber hoses you’d have
to replace!! Not to mention rubber tires and rubber toys!

Yes, it was the same unit. He did not try to sneak in his
friend’s for repair. I started asking him what he could have in
his house that could do this. It turned out it was the electronic
air cleaner. Soon after that, a lady brought in a six-month-old
VCR and the pinch roller looked 20 years old. “You have an
electronic air cleaner,” I told her. She looked a little shocked
and asked, “How do you know?” After some thought she said,
“Now I know why the rubber on my patio doors is gone.”

Then there was another customer who turned white when I
told him about air cleaners. He stated, “I have to replace all my
windows at a cost of $4000.” He went on to tell me about how
he had just replaced his carpet as the rubber backing had
turned to dust.

• Jonn van Laar (via e-mail)

• Pease: Hi, John. Wow ; I’m impressed. Good sleuthing. Thanks
for the info. We’re publishing your warning. Please tell me
more about the “electronic air cleaners.” Is this one of those
types that emits negative ions? Please find out from your cus¬
tomers the name of the maker and the model number. How is
it described? Note: when the rubber on the patio doors is
replaced (and the rubber mat under the rug), there may be
some synthetic material, such as neoprene, that will resist the
ions a lot better than rubber. I’m not much of a chemist, but
your customers might want to contact one!!

Jonn van Laar responds: Hi, Bob. I found the “Living Alpine
Fresh Air Purifier Model 3500HL,” also known as “Healthy Liv¬
ing HL-2,” on eBay. It seems similar to the one that was alleged¬
ly causing the trouble. (Yes, that “electronic air cleaner” is the
type that costs several hundred dollars, and it emits negative
ions and "activated oxygen”—ozone. I guess that is the main
problem!!/rap)

We had a guy selling them for as much as $600.1 remember
smelling the air and it was just like what you got in the old days
when you held a screwdriver close to the plate cap of the hori¬
zontal output tube on a TV. So my guess is that it is ozone.

What got me was that it could do this to a belt inside a TV/VCR
that is not open to the air—and in less than a month! (Hey, that
activated oxygen really does get in everywhere!/rap) I do not
know what kind of rubber it was. I am sure different kinds
would react in different ways.

The unit uses some kind of RF signal because I remember look¬
ing on a scope. Just holding the probe close showed all kinds of

Dear Bob: Mr. Ed Ganshirt wrote to you about a 120-A rating
on a T0-220 packaged MOSFET (electronic design, Aug. 18, p.
22). I think that this must have been a misprint (or misread) of
what was meant to be 120 W.

• Kent Nickerson (via e-mail)

• Pease: You can go look up the datasheet, and it is very
clearly amps, not watts. Go ahead and Google the IRF1407-
ND! Besides, a TO-220 definitely can not handle 120 W, not
even with an infinite heatsink!!

HELP! PLL QUESTION • I got an e-mail from a guy inquiring how
to make a precise 59-Hz signal using a phase-locked loop (PLL),
and my hateful computer ate the message before I could reply.
Amigo, please resend your message, as I do know howto solve
your problem. It’s easy, using Application Note 210. /rap GO

ED Online 11084

Comments invited! rap@galaxy.nsc.com —or:
Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, Calif.

• 18 •

Gotowww.elecdesign.com 09.29.05 Electronic Design

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHAT'S ALL THIS MARKETING
STUFF, ANYHOW?

Okay, marketing is a dirty and nasty job, but somebody has to do
it. Sometimes I put on my “marketing hat” and try to do some
of it myself. But the best insight into that is the quote I have
paraded before a couple thousand attendees at my recent
linear seminars: “The only valid market survey is... a signed

purchase order.” Well over 2000 people attended these semi¬
nars, and nobody knew who said that. Answer below. Hint: He
was one of the Fairchild Eight. Your chances of guessing the
right answer are about one in 10.

Marketing has several tasks, all involving judgement (i.e.
guessing). The marketing manager has to come up with the
answer to: “If we brought out THIS product, with THAT set of
features, would it be popular?”

If you guess right, you get to try again. If you guess wrong,
you get to try again—at a different company.

Here at NSC, many people have tried to guess what the mar¬
ket would favor. I have heard several contradictory stories of
which products were recommended by whom. Nobody ever
said it was simple.

When I was at Philbrick, we had lots of great and profitable
products that were invented by engineers. I guess there were
some marketing guys, but I don’t recall who. But when we had a
great new product idea, they got the heck out of the way, greased
the slides, and let us bring it out. Bob Maker’s P2 was a fantastic
example. Nobody ever asked for it until it popped out full-fledged.
My 4701 series comprised very popular V-to-F converters, and
nobody ever asked for them.

When I was leaving Philbrick in 1976,1 looked at several oth¬
er companies. Some were run by engineers, and some were
run by marketing people. The companies run by marketing peo¬
ple seemed to be proud of how they could imagine a new prod¬
uct that was great and force the engineers to make it. I could
see where they had forced some amazing but overpriced (and
not very attractive) products into the marketplace. But they had
failed to bring out some naturally winning products. I recall that
after I’d designed those V-to-F converters, the marketing guys
at one company refused to believe there was any “market for
VFCs.” But finally the sales guys and engineers convinced
them that while there was no “market” for them, there were
customers, so they finally got into that business.

When I joined National, it was the other way around. In those
days, it was easy to invent a new product that people would
like—and love. We didn’t have to ask the permission of any mar¬
keting manager. We just had to put on our marketing hats. It
wasn’t really hard, and some of us got pretty good at it.

I mean, every new product I designed started with the con¬
cept of a datasheet for the new product. I would invent all sorts
of features I wanted to make the datasheet look great. Then I
would start my circuit design to try to meet all those specs and
features. Occasionally, I had to redesign the datasheet, or the
circuit, to get closer agreement. And all this time, I kept think¬
ing about the features that would make me happy if I were a
potential customer. Thinking like a customer is not too bad a
way to design products.

So the guy who said that quotable saying was Jay Last, one
of the Fairchild Eight and a very knowledgeable guy. In his con¬
text, it was largely true. His marketing guy could ask many cus¬
tomers, “Who would like to buy THIS integrated function, or a
new silicon planar THAT?” Very few people would say “no.” But
did they vote with their dollars when the product came out?

Not so fast! OO ED Online 11138

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
RO. Box 58090 , Santa Clara , CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

... 20

10 . 13.05 Electronic Design

• •

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB’S MAILBOX

Hi Bob: I've been an avid fan of yours for many years, and I'm writing now to plumb the depths of your arcane knowledge. We've got a
design that needs a little bit of insulation, and I told the engineer to use fish paper to do the job. She asked me a question that's got me

stumped: "What's the origin of the term fish paper (and how did it come to be associated with vulcanized fiber)?" A search on the Internet

left me still scratching my head. Any ideas?

■ Matthew L. Severns (via e-mail)

■ Pease: If you go to www.jjorly.com/fish_paper_die_cut.htm, it will explain about fish paper—why it has an electrical spec
drawing and conforms to MIL-I-695F, and why it is rated to high temps with good electrical and mechanical characteristics. I
gotta admit, it does not seem to explain the source of the name. On the other hand, www.smallparts.com/products/descriptions
/fshp.cfm says that fish paper is made out of cotton rags. Maybe it's thus named because it smells like fish. Does anyone know the
origin of the name?

Dear Bob: I want to raise a concern about the gaining popularity of using spread-spectrum techniques in switching power supplies (As in
other digital computing chips... /rap) to allegedly reduce EMI. Of course, this does not really reduce the EMI. It just spreads it out so the
equipment can pass the regulatory tests, which are performed using narrow-band detectors. Because a real receiver like a TV set is
wideband, in practice, the interference is not reduced. (The total radiated power may actually be increased, as there is no need for
shielding or good layouts, /rap)

One advertisement shows a CW interference carrier being reduced by what looks like an impressive 20 dB. But the reduction is only
because the measurement is being made in a 100-Hz bandwidth. With the proliferation of wireless RF technology, I think we all need to
design products that truly pass EMI requirements without resorting to tricks that exploit loopholes in the test methods. With the
proliferation of switching power supplies and unfiltered lamp dimmers, AM radio has already become almost unusable.

■ Mark Kolber (via e-mail)

■ Pease: Well, every loophole will be investigated, and if they seem to work legally and save money, people will use them. Hey,
shielding is expensive! And making a good new re-layout usually takes a lot of time—and it's hard to be sure that it will work! I
don't see the FCC cracking down on this in the near future. Heck, people are permitted to spray broadband noise over our power
lines (and they are not even twisted pairs)! Permissiveness seems much more rampant. Regulators don't give a damn. Anything
that lets you get out a product faster and cheaper is a great idea, no?

Dear Bob: I am remiss not to write sooner—been terribly busy with hurricanes, grandkids, work, etc. I just got to your column on naps
(ELECTRONIC DESIGN, May 12, p. 20) and I found it to be very funny, but quite accurate. Could it be that you and I are getting a bit long
in the tooth? I am now 73—the days of working with Bob Widlar, Brent Welling, Brian Hollins, and Bob Swanson are a faint, but fond,
memory. (I was with you at National Semi from 1976 to 1980.)

As I have told you, I still work full-time consulting for the USAF and plan to continue to do so as long as my health will allow. But on the
subject of naps, I find that going home during the lunch break and taking a 15-minute "power" nap will refresh me entirely for the
remainder of the afternoon. I have no trouble working a nine-hour day (with a 45-minute lunch period). So, I totally agree with you—naps
are good things. And I really enjoyed your column.

■ Jerry Robertson (via e-mail)

■ Pease: Surprisingly, I usually get through the day without a nap during lunchtime. Not sure why. Keep up the good work!! Best

These transformers boosted my toaster from 1260 to 1460 W.

October 27, 2005 Electronic Design

• •

J Pease PORRIDGE

t BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHAT’S ALL THIS LOGARITHMIC
STUFF, ANYHOW? (PART 2)

The l c versus V BE of modern transistors has some excellent log
characteristics, as we have discussed."' If you ground a transis¬
tor’s base and compensate for its Vbe with a matching Vbe?
you can do some good logging over a wide range, from 1 mA
to 1 pA, and really quite accurate from 100 pA to 100 pA.

That’s six decades where the limitation of R EE ’ on the high end
is the major limitation. Input leakage currents on the low end are
not really a limitation, these days, as good CMOS amplifiers have
l !N smaller than 1/1000 of the 100-pA signal. Of
course, this works best around room temperature.

The compensating resistor R x (1-kQ wirewound
resistor at +3500 ppm/°C) works fairly well for a
moderate temperature range, and it works best
around room temperature. Figure 1 shows a stan¬
dard log ratio circuit, found in AN-29 and in many
books. Its output is -1V per decade for inputs larg¬
er than 0.2 pA.

However, there are definite limitations in speed
of response. A large input current requires a large
feedback capacitor, such as 500 pF, because the
transistor adds so much gain to the loop that it
ruins the loop stability. When l jN is as large as 1 mA, the transistor
has a gain of 800—too much gain to add to an op amp. (Op amps
are happy with attenuation in the feedback loop, but they don’t
like gain added.) So we have to add a large C F , such as 500 pF, to
make the loop stable at high frequencies.

Now when l )N is decreased down to 1 pA, the loop is very stable
and very slow. The 500 pF is much too big. The bandwidth falls
below 1 kHz. A feedback capacitance of 2 pF would be plenty.
Some of our customers needed a logger with good audio band¬
width over a wide range of input currents. What to do?

I remembered that some engineers back in the 1960s had this
problem and used some diodes to solve it. But I never saw how
they did it. So I went back to the scene of the crime. All I needed
was a feedback capacitance that was big when the signal was
big, and small when the signal was small.

I figured out that the circuit of Figure 2 might do it. All I had
to do was build it up and try it out. Do a cut-and-fit on the
capacitance sizes. Sure enough, it worked quite well. The effec¬
tive feedback capacitance increases as the output goes more
negative, and the diodes start to conduct. I used a fairly spa¬
cious layout, so the stray feedback capacitance (C 0 ) was only 1
or 2 pF, due mostly to the amplifier’s socket and the 1N914. If
you made a really good layout, you could get the strays even
lower than that and get fast response below 1/10 pA.

I kept a bandwidth of 12 kHz, from 1/4 pA up to 0.4 mA. I did¬
n’t really try hard to optimize it further. I stopped, as it was work¬
ing entirely well enough for me. It did overshoot a bit, but it did not

ring at any signal lev¬
el. You have my per¬
mission to optimize it
as needed. You
might use three or
four diode-R-C net¬
works to cover a
wider range or to get
cleaner response at
all levels. So there’s
an old trick that has¬
n’t been seen for about 40 years, but when we have to recon¬
struct it, it works pretty well. £0 Onlm® UWt

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090 , Santa Clara , CA 95052-8090

* “What’s All This Logarithmic Stuff, Anyhow?” Electronic
Design, June 14, 1999, p. Ill; ED Online 6068 at
www. elecdesign. com

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
Nationai Semiconductor Corp., Santa Clara, Calif.

vV <2.oy:

-gs or^.1%

= 1S3K

±VS = -f24-3.^
ffj )• 1/ 3h&\ C- LoJJ <? r

Co, s«.« -text

AXp-nV*- Capacitance.

.. 20 • •

11 . 07.05 Electronic Design

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

BOB’S MAILBOX

Dear Bob: I had to laugh at your recent column. 51 ' I experienced
similar product destruction with my (brand x) air cleaner. Let’s
see, I lost: all belts and wheels in two high-end (industrial)

VCRs —and one a second time,
players, in a DVD player, three

DVD drive, plus a belt in a vacuum cleaner—three times.

(Ouch!! That’s painful!!/rap) The air cleaner seems to be two
appliances in one. It has an ozone generator and a positive ion
generator. The positive ions don’t destroy anything but make
your walls and windows dirty by depositing dirt on them. If I
were the litigious type, I’d smell class action.

• John Spangler (via e-mail)

• Pease: Well, l can’t say I’m in your class action, but I am try¬
ing to help by warning everybody. Note that some compa¬
nies, such as the Sharper Image, will sell you an air cleaner
guaranteed free of ozone.

Dear Bob: I was the chief engineer of a Cleveland, Ohio, FM
radio station back in the mid-1970s. Inside the transmitter
building was the usual high level of ambient noise from the
transmitter blowers. To hear program audio I had a fairly hefty
receiver and speaker system. One day I noticed that there was
something not quite right about the sound.

Pulling the front cover off revealed that the compliant rubber
ring supporting the cone had completely disintegrated on both
units. (Uh, yeah, /rap) I never actually smelled any ozone, but
I’m certain some was being generated by the high potentials.

Judging from the September issue article, there must be
some pretty elevated levels of ozone in those living spaces to
cause the problems mentioned. Sounds like we should be
more worried about ozone than radon gas!

• John Matteslni (via e-mail)

• Pease: Yeah, but a transmitter has a set of big fans to keep
it reasonably cool. And the hotair is then sent out of the
building, along with most of the ozone. So it’s not as bad
there as in some other places.

Hi Bob: I certainly do agree with your comments in “What’s All
This Wireless Stuff, Anyhow?" (electronic design, Sept 15, p.

22). Pictures are indeed better on the radio. A few years ago,
Stan Freeberg, a once-notable comedian, made a series of
commercials for radio, which were broadcast on (what else?)
the radio. The purpose of the advertisements was to vividly
illustrate the power of the human mind to visualize objects and
tasks that would otherwise be impossible to portray on televi¬
sion or other visual media.

Imagine, as did Stan Freeberg, Lake Erie totally covered in
whipped cream with giant battleships delivering maraschino

Also, belts in three audio CD
computer CD drives and one

cherries! I can actually “see" that in my mind, can’t you? (Heck,
I got it memorized! It’s a classic. But wasn’t it a bomber that
dropped the cherry on top? I can still see that in my mind’s eye.
I don’t recall seeing any battleships./rap)

The “eye of the mind" has always been more graphic and
demonstrable than any graphic that can be presented.
Remember all those radio programs that you and I listened to
as kids? How did the “Green Hornet” look to you? How about
the squeaking door on The Inner Sanctum?

• Karl Kanalz (via e-mail)

• Pease: Man, Inner Sanctum was too scary for me. It made
me uncomfortable. These shows were a little too real!

Dear Bob: Your article on “Wireless Stuff” was most amusing.
On a similar note, you might want to look at the wired world as
well. In the 19th century, we had cable communications. And
the system was totally digital. It was called the telegraph.

• John A . Rupkalvis (via e-mail)

• Pease: Gee, I don’t even have digital cable communications
at my house. I do have a telephone, but it’s analog. ©©

Comments invited! rap@galaxy.nsc.com —or:
Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

* ELECTRONIC DESIGN, Sept. 29, p. 18

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, Calif.

.. 18 • •

11 . 17.05 Electronic Design

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHAT’S ALL THIS AMT STUFF,
ANYHOW? (PART 1)

The IRS recently “invited” me to fill out a form that puts fear in the
hearts of strong men—Form 6251 for the Alternative Minimum
Tax (AMT). I was “invited” to find out if this form would show
any more tax liability. Hey, I paid my taxes back in August. I
don’t mind paying my taxes. I don’t mind filling in my tax forms.

I have always done my taxes by hand
without a computer or any profession¬
al preparer. (Actually, I never have any
trouble filling in my tax forms. It's just
collecting the numbers to plug in
that’s a lot of work.)

So I started trying to fill in this form.

Unlike any other tax forms, it refers in
a circular way to other tax forms and
worksheets that have supposedly
been completed in the past. Unfortu¬
nately, some of these alternative work¬
sheets are hard to dig up, so I had to
re-duplicate and re-invent them. The
lines to fill in had a lot of circular think¬
ing. I had to refer to various lines on
worksheets for other lines. Often I was
told to add zero to zero, then compare
this to another number, possibly also
zero. Then I had to multiply by 15%,

28%, or maybe 26%. It did seem to be
a rotten and absurd form.

The AMT was originally designed in
1970 as a tool to harass tax-avoiders. Thirty-six years ago, some
citizens and legislators were outraged to learn that 155 people
with over $200,000 dollars of income (which was a lot of money
in 1969) were paying zero income tax. They did this by exercis¬
ing various deductions and exemptions and tax loopholes. So
the AMT’s purpose was to make them pay something .

Now, 30 years later, outraged people are complaining that
there are still over 600 millionaires who are paying nothing in
taxes. They have an income of over $1M but are still not paying
income tax or even AMT. However, because it was not indexed
against inflation, many more ordinary people and taxpayers are
creeping up into the -$100,000 brackets where they (we) get
nailed by the AMT.

Apparently you get into the AMT by having lots of deductions,
such as state income taxes, mortgage interest, various kinds of
incentive stock options, and exemptions for children. When Con¬
gress passed our tax codes, imperfect as they are, they wanted

Ito provide incentives for people with
ithese expenses to have a tax break.
iBut the AMT has no such incentives. It
mails you on those items.

I had computed my AMT back in
f August and was satisfied that I owed
$000 of AMT—as in previous years.
After all, I didn’t have any exemptions
for dependents or incentive stock
options. So on Oct. 23,1 started re¬
computing my AMT. I was quite sur¬
prised to find that I supposedly owed
$5414.1 didn’t like that answer, so I
re-computed it. Maybe I had made
some errors. In that buggy, corrupt
AMT form, it would have been aston¬
ishing if I hadn't made an error. I
cranked through it on clean sheets of
paper and came up with $17,300.1
checked the first one and the second
one, and I could not find any obvious
errors. I started from scratch and
came up with an answer of $4222. I
compared this computation with the previous two, and I could
now see where I'd probably made errors.

I decided to check it again, and it seemed to tell me $3878.
But I was nervous about that. I sent in the form for $4222. If
they tell me I made an error, and it's really $3878 (plus penal¬
ty) , I won’t be surprised. I won’t be surprised at anything.

If you have ever been surprised by an AMT tax bill, check
550 on the Reader Response card. If you think the Form 6251
is the worst form of all, and worse than all other tax forms put
together, check 551.©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052 8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

• * 20 • •

12 . 01.05 Electronic Design

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHAT'S ALL THIS AMT STUFF,
ANYHOW? (PART 2)

As I said in my last column, I don’t mind filling in my tax forms. 51 '
I can usually look at the 1040 forms and figure what is happen¬
ing. If I am normally in a 33% bracket, I can tell that the tax is
really at a marginal (incremental) rate of 36.8%, or whatever.
But in the AMT, I cannot decipher my marginal rate.

I may get the nerve to go
in and increment my income
by $1000 and see what the
real change is. But it would
take me a couple of hours,
and I’d have to be sure I was
not making any mistakes in
the first place. I’m guessing
it’s up near 38.9%, due to
discontinuance of various
exemptions.

The big trouble with the
AMT is that it brings in so
much money! Many con¬
gressmen have expressed
their unhappiness with the
AMT, but they recognize that
if they want to reform it, they
will have to find some other
source of funding for the
government. And, (R) or (D),
nobody has a good idea how to do that.

I think I know what I’ll do next year. I’ll get my taxes all done in
July, the best I can do it. Then, I’ll turn in an advance copy to a
good honest tax preparer and get his opinion if I have done
everything right. If I’ve done anything wrong, I’ll leave time to
recompute and recopy them before August 15.1 wouldn’t trust
the preparer to do them for me, but I’ll look at what he has to say.

Is there any way to plan ahead and avoid paying AMT? Not
very easily. There are discretionary expenses you can take in
different times to minimize your conventional 1040 taxes. But
that doesn’t work well for AMT.

One thing you can do is avoid exercising a big stock option and
getting hurt if the stock goes way down. I wouldn’t mind exercising
a stock option and having to pay taxes on my gains. But if I exer¬
cise and the stock goes down, 1 will have to pay AMT on the stock
gain, even if there was a loss making the stock basically worth¬
less. If I take an option, I always sell some of it, to make sure I
don’t take a beating, paying taxes on something that’s worthless.
Usually I prefer to sell old stock, of the same type, so I only have to

pay taxes on long-term capital gains. At first
when I started studying the AMT, I thought I
would get screwed out of the 15% rate for
long-term capital gains, but apparently not.

So when will Congress do something
about this AMT injustice? When enough
people complain. Every year, more and
more people are caught up in this net, and
when enough complain, maybe we’ll get
some action. And if I sell this column 10
times over, maybe it will pay for my AMT bill.
No, not really.

BOB’S MAILBOX

Hi Bob: ...I would like to know how you nor¬
mally do the following: In some high-fre¬
quency circuits, for instance in resonant dc-
dc converters operating in the range of a
few megahertz, it is necessary to look at
current and voltage waveforms simultane¬
ously. How do you ensure that the current
and voltage waveforms have the same time delays so they can
be compared?

• Ja vier Chivite (via e-m a il)

* Pease; / don ’t use a current probe very much , but l guess you
just put in a calibration test For example, if you have a nice crisp
voltage, put that on one probe, add a 1-kQ resistor to ground,
and let the current probe see that current. If there is any delay,
such as a few nanoseconds, you can easily see it and subtract it
from ongoing measurements. Write a note on the current probe
to help you remember it 0©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

* ELECTRONIC DESIGN, Dec. 1, p. 20

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

. . is ..

12 . 15.05 Electronic Design

• •

What’s All This Future
Stuff, Anyhow?

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

I was down at an “Arrowfest” in Plano, Texas, a few months
ago. We had a little panel session about the future with Bill
IQein of Texas Instruments, Arnold Williams from Analog
Devices, and myself. Most of the audience agreed with the three

of us (and the moderator)—the future will have a lot of analog.

When we see that the power-supply drain of a fast processor
at 0.09 pm is largely related to the device leakages—which
aren’t very predictable and can’t be modeled easily—hey, that
processor is relying on analog characteristics.

Remember the old days, when engineers were struggling
to bias up germanium transistors so their bias wouldn’t
suffer from thermal runaway? Now, the silicon circuits
are on the verge of thermal runaway.

It will take some good, tough engineering to solve
the layout and heating problems. Maybe when
they turn off the clocks to some subsys¬
tems, that area of the chip will stop self¬
heating, the temperature will go down,
and the leakage will decrease—a
kind of thermal un-runaway.

I wonder what Moore’s Law
has to say about that. Obviously,
when the “process” goes down
to 0.065 pm, it will get much
worse. Now that designers
can’t keep shrinking things, the
digital field is going to look a lot dif¬
ferent. What will they sell us next
year? Microprocessors that aren’t just
fast battery dischargers?

FUTURE POWER • People are still bringing up the
old saying that “Fuel cells are the power source of the
future—and always will be.” Down here in Silicon Valley, the
Transit Authority operates a couple of busses running on fuel cells.
This is an ideal area, because it’s so flat with very few hills. (They’d
never cut it in San Francisco.) From what I hear, the fuel cells have
a very finite life and will need to be replaced every few months. And
they’re expensive! I’m waiting to hear how those busses work out

Engineers, scientists, and chemists have been trying for
many years to reduce the price of fuel cells by an order of
magnitude or two. Great! NASA is very happy to pay for such
exotic energy sources. Now the fuel-cell makers only have to
improve the cells by three or four more orders of magnitude
before they fit into future cars.

Are high-efficiency solar panels headed for a big future?
I’ve seen studies that say inexpensive (not-so-high-efficien-
cy) solar arrays are the right way to make cost-effective
energy. It’s fine by me, if the cost of a watt keeps going
down. Mr. Ovshinsky of Energy Conversion Devices (ECD)
has been saying that for years, and people are starting
to follow him.

Furthermore, people are beginning to see that nuclear
power plants have a future. It’s better than pretending
we can buy cheap oil forever. Besides, nuclear
power has a much more reliable availability
factor than solar power or wind power.

All we have to do is get the politicians
to agree to hide the nuclear spoils
where they won’t contaminate us
and poison us all.

And maybe this time, we
can make sure the shale-oil
conversion experts get
going and keep going. The
U.S, has enough oil-shale to
last us many decades into the
future for a large fraction of our
energy needs. We only have to
prove that we can rely on it at
some plausible, reasonable price—
and not chop the support for that
energy when the oil shock decreases.
So while I’m not very good at pre¬
dicting the future, lots of other people
are predicting good things! I just hope I’ll be hanging
around for many more years to smile at them when they
get here! Nuclear-powered op amps, anyone? ©

ED ONLINE 11688

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE brom MIT in 1961 and is Staff Scientist at
National Semiconductor Corp-, Santa Clara, California

r iE-»

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

Bob's Mailbox

H i Bob: Back in the early ’70s, I was an EE major at
Lowell Tech in Lowell, Mass. Within easy walking
distance of the dorms, there was an electronics
surplus store in one of the old textile mill buildings.
One day I found some Philbrick P65AU op-amp modules there.

I bought several of them, for maybe $5.00 total. (A bargain,
/rap) I decided to build a stereo hi-fi amp around them. It took
the better part of the semester to build it, and I nearly flunked
out as a result. But I ended up with a pretty good stereo amp at
the end. I learned a lot by building the amp, a lot more than
they were teaching me in class at the time!

Limited by the 30-V supply the op amp would run at, it didn’t
put out a lot of power. But with some efficient bass-reflex
speakers, it was loud enough. (Check. The P66s were useful? If
you had enough of them, you could parallelseveral of them or
add other boost transistors, /rap)

A few years later I took the amp to a Tech HIFI audio clinic,
where they got a good chuckle when they saw my homemade
amp. But when they ran the frequency-response and distortion
checks, they were quite impressed.

(The P65 wasn't very fast for slew-rate. It had about 0.6 V/ps,
so at 20 kHz, it could barely swing 8 V p-p. But most program
material doesn't have a large amount of content at 20 kHz. The
P65 also was a good low-noise preamp for phono cartridges,
etc. /rap)

I used that stereo amp for many years and eventually gave it
to a friend of mine, who might still be using it. I still have a
“spare" P65AU module and a P66A booster follower, which I
never did find a use for.

• Steven Weber (via e-mail)

• Pease: Take good care of them, Steven! We may need them
someday! I’ve recently been running some P65s in a 1-ms
delay line, using 10 H of inductance.

HI Bob: My son is a chemist, so I sent him an e-mail asking him
about the ozone reaction on rubber and rubber products (elec¬
tronic design, Sept. 29, p. 18.). Here is his repjy:

“Ozone is a highly reactive form of oxygen, which is why it
can be dangerous to human health as well. Ozone is one of the
primary causes of the degradation of rubber because it reacts
with any unsaturated double bonds left in the rubber polymer.

It can also react with regular covalent-type polymer bonds, but
this is more of a minor reaction. Once the reaction f\as started,
it basically goes down the polymer chain via free-radicals and
‘unzips’ the polymer molecule—effectively destroying it. Tire
manufacturers minimize the effect of ozone degradation by
adding antioxidants to their tire formulas and waxes that pro¬
tect by blooming to the surface of the tire and forming a coat¬
ing, effectively sealing the surface molecules of the rubber
from the air and ozone. This is why the tire manufacturers dis¬
courage the use of products like Armor All on the sidewalls of

the tires because the Armor All can solubilize the waxes and
then remove them overtime.”

• Dave Miller (via e-mail)

• Pease: Hello, Dave. Well, that certainly seems to be the
right story on ozone. If it beats up rubber, imagine what it
does to our covalent bonds! That's what they were com¬
plaining about when Los Angeles was so smoggy! It's inter¬
esting how tire makers protect their tires. Too bad rubber
belt makers can't do that Or our lungs!

Dear RAP: Since you were quoting only the lyrics from “They All
Laughed” (electronic design, Sept 15, p. 22.), and not the
music, perhaps the credit should have gone to fra Gershwin,
instead of George. George just did the music...

• Bruce Walker (via e-mail)

• Pease ; Hello, Bruce. Well, you are quite right. I must have
goofed. My big reference book, American Popular Song by
David Ewen, did say exactly that I'm sorry! ©

ED ONLINE 11865

Comments invited! rap@gaiaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, Calif.

18 *“

01.19.06 ELECTRONIC DESIGN

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

What's Al This Crampon
Stuff, Anyhow?

B ack in February 1955, when I was just 14,1 snowshoed
with Professor John A. Williams, “Alpine Fred” Tor-
rey, and a group of Explorer Scouts up to Crag Cabin
at the 4200-ft level on Mt. Adams in New Hampshire.

I was strong enough to carry a little extra weight, so they volun¬
teered me to carry some crampons—a set of spikes for climbing on
ice and hard-packed snow. I carried them, though I had no idea
what they were good for.

After we ascended up to 4400 ft, we started onto some ice and
frozen snow slabs. Wearing the crampons, I immediately realized
that I was master of the snowscape. After I helped escort my
friends over the icy spots, I had the best fun hiking. I could amble
easily across these slanted drifts, at any angle, without worrying
about the rough rocks or lumpy shrubs under the snow. It made
ascending the 5798-ft mountain easy and fun.

The next February, I decided to lead my own trek up there. I
was just 15, and a good camper and hiker, and I knew how to
do all the things we had done a year earlier. So we planned
carefully and got all the necessary clothes and equipment. It
never got below -20°F, and we had a great hike. All the hikers
had crampons this time, and we strode easily across the frozen
snow, farther than ever. In later years, we went back for more
of such climbing, and it was wonderful.

Now, 50 years later, we’re going to climb up there again for a
50th anniversary trek. But one of my friends has very large boots,
and it’s hard to find crampons to fit them. What to do?

In November, I was trekking up in the Annapurna region of
Nepal. (See my trip report at www.national.com/rap.) I had brought

n*i

along some ice creepers, in case we were waylaid by a lot of ice or
snow. I tried some tests up above 14,000 ft at Annapurna Base
Camp and walked across some frozen snow. The four-corner
stamped-steel ice creepers worked, with traction just a little better
than good climbing boots (when they didn’t fall off).

Likewise, I evaluated some Yaktrax, and they also provided good
traction. “If I were crossing an icy parking lot, they would work just a
little better than a few handfuls of salt or sand,” as I explained to
my buddies.

“But there are no parking lots up here,” they replied.

“Sure, this hard-packed snow and ice over here is just like a
stomped-down parking lot!” I said. I motioned to a strapping young
Nepali porter, and he put the creepers on. We started a friendly
pushing contest. Both types had surprisingly good traction. There
was no serious advantage. So I found that the Yaktrax are suitable
for icy or slippery, snowy parking Igts. The four-corner steel cleats
weren’t quite as good, but they would save you from falls on a flat
area of slippery snow or ice.

Later I went shopping in Kathmandu and visited the used-equip¬
ment markets. I found several kinds of excellent new crampons,
but at good, high prices. I mean, $120 to $180 is a fair price if
you’re going to do technical ice climbing, but I wouldn’t pay that
much. Finally I found the secondhand ones shown in the photo,
adjustable up to 14 in., with excellent welding, for just $21. So we
are all set for our New Hampshire trek.

Would I bring the Yaktrax as backup for our good crampons?
Nah. But I will bring some baling wire for any needed repairs. The
main problem with crampons (assuming they don’t break) is that
you can easily slash the legs of your pants-or your legs-with the
sharp spikes. We plan to hike slowly and carefully and use lots of
rubber bands to keep the legs of our pants from flapping. ©

ed tmum.

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara , CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

02.02.06 ELECTRONIC DESIGN

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

Bob's Mailbox

ear Bob: I have a little microcontroller circuit pow-
I M ered from a 115-V ac to 24-V dc supply that is then
■ Vfurther stepped down to 5 V with a buck converter.
An RS-232 converter, whose switched capacitors
give ±8-V swing, is used to communicate to a PC. The problem
I have is that connecting to a grounded PC has blown out the
serial port on the PC. We grounded the common in the circuit
and have had no problems since.

I’m looking to understand what possibly could have hap¬
pened. Could the 24-V supply have floated up and accumulat¬
ed enough charge to blow out the port? If so, how does this
happen? I have also seen in many circuits a 1-WIQ. or larger
resistor and a 0.01-p.F capacitor in parallel tied between the
power-supply common and earth ground. Is there some stan¬
dard or something else I missed that suggests this? Can you
please enlighten me or point me to a reference? Thanks.

• JeffSIefke (via e-mail)

• Pease: lam not an expert on this, but I can make a guess.

(A) The ground of the RS-232's supply must have floated up
to a high voltage, perhaps more than 100 V. There probably
was something like a 0.01-jiF capacitor, as you suggested.

(B) When the pins of the connector were connected to the
serial port, the ground pin (or pins) did not connect first. (If it
had, it would have resulted in no harm.)

(C) In a well-designed system, the ground pin (or pins) would
be longer than the signal pins, so it couldn't help but con¬
nect first. If the signal pin connected first, that might cause
some harm. And it did.

(D) A well-designed serial port ought to have 1 kD in series
with the inputs, going over to some protect diodes. But
these days, everybody leaves protection components out of
their computers because “ it's too expensive" or because

u extra components hurt reliability." That sounds like enough
to explain it.

HI Bob: I was teaching a unit on transistor amplifiers yester¬
day, and a student asked me why transistors are labeled as
Ql, Q2, Q3... in schematic diagrams. So as an old timer with a
wealth of trivia, why the letter Q for transistors? Love your Elec¬
tronic Design column.

• Michael Halbern (via e-mail)

• Pease: When vacuum-tube circuits were developed, all the
other letters were used to identify Transformers, (Vacuum)
Tubes, Diodes, Rs, Cs, Ls—and Q was left over for " miscella¬
neous. " When transistors came along, they were about as
miscellaneous as you could get, so they became Qs.

Hello Bob: Last night I was replacing the CR2032 battery in
an inexpensive digital tire-pressure gauge, when the small liquid-
crystal display fell out. On closer inspection, I found it had no
wires or electrical contacts whatsoever. I was able to sandwich

the LCD and two associated foam pieces back between the PC
board and the housing, and it worked. How do they do that?

• Steve Goss (via e-mail)

• Pease: The foam is made of alternate vertical stripes of con¬
ductive and insulating foam. Both are kinda squishy. The
pitch of these is like 10 times smaller than the pitch of the
electrodes on the PC board and the LCD. And I bet you can
see these, presuming you look real close.

Dear Bob: Read with interest your experiences and recom¬
mendations on napping (electronic design, May 12, 2005, p.
20). When I am caught by my boss with head on desk later this
afternoon, I shall be sure to have on hand your invaluable
guide to use in self-defense.

• Howard Jones (via e-mail)

• Pease: No, all you have to do is lift up your head, lift up your
eyes, lower your eyes, and say, *Amen." Then stand up and
ask your boss what you can do for him. Of course, this does¬
n't work very well if you're snoring! ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara , California.

18 -

02 . 16.06 ELECTRONIC DESIGN

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

What's All This Autosave
Stuff, Anyhow?

hen I first began to type on a computerized word¬
processing system, I realized that a power outage
was potentially disastrous. You could type a care¬
fully crafted four-page memo, nicely polished, and

it all would go away if the power
went out—even for a second, or
if somebody kicked the cord a
little too hard.

When I used the Coleco Adam,
the process of saving was quite a
big deal. Copying onto Adam’s
audiotape took well over two min¬
utes, and you could prepare and
drink a cup of coffee while wait¬
ing. Worse yet, if you saved the
text too often, you could fill up
one tape with 14 saves. Each
document “save" was saved sep¬
arately and didn’t write over the
previous save. This was okay, bet¬
ter than no save at all, but not great.

Lately, l have been getting fewer and fewer power outages. A
laptop with its own battery power is substantially immune from
power loss, too. But I can still lose my story if the computer
locks up. Just last week, I typed for an hour, and the computer
locked up and refused to let me save or copy anything. Thanks,
Bill. Time to retype from memory.

In between, I was using PC Write Lite from Quicksoft (no
longer in business, alas). One of its nice features was its
autosave. I could just set the autosave for “10 minutes min or
2000 bytes min." After that amount of typing, it would stop
briefly and save what I had typed. I consider that to be one of
the finest features of a civilized computer.

When my wife got a Mac, I asked her how her autosave worked.
Her reply was, “the what?" I looked in her computer manuals, and I
was horrified to learn that there was none. I looked in the big, com¬
prehensive book on OS-X by Robin Williams. The best advice she
has is “SOS,” that is, “Save Often, Sweetie.” Thanks a lot, Robin.

There are too many ways for computers to goof up, destroy
your carefully typed messages, and leave you in the lurch. I’m not
a connoisseur of error messages, but I’ve been shafted in about
80 different ways that lock up the computer or destroy a mes¬
sage. Sometimes there isn’t even any error message. It’s just all

locked up. So maybe I should get
a cheap little timer, set it for 10
minutes, and save every so often.
For sure, five minutes of lost typ¬
ing is a lot better than 50.

Somebody told me that
Microsoft Word has an autosave.
You can set it for every so many
minutes, but not for the number of
bytes. I guess that would be better
than nothing, but a fast typist still
could get in trouble.

Recently, I had a couple of e-
mails arrive about three days
late. One of them was blamed
on a balky firewall, which has
pro'Daoiy oeen fixed.Tne ot'ner was blamed on some balky
aspects of Mozilla 1.0. Maybe the newer 1.5 version will fix
that. Usually, only spam is delayed. But this time I noticed
that a colleague had sent an e-mail that arrived three days
late. I sure noticed that discrepancy.

I’ve had other problems recently, too. A couple of e-mails
arrived and then disappeared. A boffin is trying to reconstruct
them from backup files, but they aren’t showing up yet. I guess
the main point is that computers hate me, and I despise them—
that is, digital computers. I get along fine with analog computers.

When I try to send myself a copy of an e-mail, the autocom-
plete guesses that when I type “rap," I’m trying to send it to my
colleague Anne Rapp. I keep erasing her name from my
address book, and it keeps popping up again. At least this one
can’t be blamed on Bill Gates.

P.S.: Yes, I have been saving this draft. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara , CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp. f Santa Clara , Calif.

20 • -

03 . 02.06 ELECTRONIC DESIGN

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

Bob's Mailbox

D ear Bob: The letter from Dave Miller* reminded
me of a recent tire episode and questions. The
front tires of my Camry needed replacing. The
dealer insisted that the new pair go on the rear
(and rotated the rear tires to the front). He claimed that
was now the recommendation of the tire manufacturers. I
recall that new tires used to be put in the front. When did
this change, and why?

(What about the recommendation of the car manufac¬
turer? Does its owner's manual give any advice? As far as
l f m concerned, recommendations of tire manufacturers
are about as valuable (and trustworthy) as you-know-what
on a boar hog. When people get stupid, / have to be very
skeptical of that advice. Of course, on a rear-wheel drive
car, you may want better traction from new tires on the
rear. Likewise, on a Camry, you may want better traction
from new tires on the front, /rap)

Also, they left a sticker on explaining that the tires were
inflated with nitrogen. What’s wrong with air, as available
at most gas stations?

• Mike Smolin (via e-mail)

• Pease: Technically, the oxygen in air can oxidize the rubber
a little. So I suppose the dealer is trying to impress a bunch
of yuppies. Or, the dealer is admitting its lousy tires are
liable to be oxidized... But I guess if we soon will be able to
refuel our cars at a hydrogen station, we can fill up our tires
at a nitrogen station. What a bunch ofcrapola. (That's a
technical term...)

Dear Bob: Just a quick
note to make you aware of
my Web site devoted to
analog electronics and his¬
tory. Check out www.ken-
nethkuhn.com/hpmuse-
um. Take the picture tour
of my shop and museum.

This may be one of the
largest home electronics
shops anywhere. All of this
vintage analog electronics
is menu driven using an
analog GUI—the instru¬
ments perform the func¬
tion the knobs point to. No
programming required. Help menus not needed. (That, I like!!
/rap) I enjoy your column in Electronic Design. Keep up the
good work.

• Ken Kuhn (via e-mail)

* Pease: Hello Ken. That's a very nice museum! At first I was
worried about those tall racks of equipment-just an invi¬

These vintage HP oscillators,

meters, and related devices repre¬
sent just a fraction of Ken Kuhn’s
collection of electronic equipment.

tation for an earthquake to knock them all down (see the
figure). But as you appear to live in Alabama, I guess they
are safe. I hope you aren't too close to that earthquake
site near New Madrid, Illinois, or Missouri, or whatever. I
guess you’re at least 100 miles away, allowing for the size
of Mississippi.

Do you have a "Wanted List" of old HP instruments or
instruction manuals? I too used to study the schematics of
the new instruments such as TEK, HP, Data Precision, etc.
These days, everything is so computerized, and most of it I
hate. Menus, I hate. I prefer your analog GUI. It's like Analog
PowerPoint. There also is a Philbrick Archive people should
check out. Take a look at www.philbrickarchive.org.

FINDING PAST COLUMNS • People often ask me how they
can find a recent column on a particular subject. It’s easy to
find most of the most recent 200 columns and Mailboxes
going back over nine years, and a few dozen before that. Go to
my Web site at www.national.com/rap. While you’re there, take
a peek at some of the horrible pictures. There are also links to
various vacuum-tube op amps, other old columns (such as Wid-
lar Stuff), and my Lists.

Next, click on the “ED Columns” and then on “click here.”
A fairly good little search engine will appear. But if you’re
interested in my stuff on “doctoring,” don’t just search on
“doctoring,” but on “doctoring stuff.” If you’re interested in
searching for recent “Ozone” topics, that turns up nicely,
even if there wasn’t any column about ozone. It searches
the readers’ letters, too. /rap ©

* electronic design, Jan. 19, p. 18

Comments invited! rap@galaxy.nsc.com —or:
Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

03 . 16.06 ELECTRONIC DESIGN

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

Whan All This Mental Math
Stuff, Anyhow?

ome night when you can’t get to sleep, try these puzzles
in your head in the dark. They’re at least as good as
counting sheep. I’ve done them in my head, and they’re
kinda fun.

1 Little Egbert bought a mountain. Its
# shape was a perfect hemisphere with a
5270-ft radius, set on a flat plain. He decided
to build a railroad to transport him to the top. It
was a monorail, which made it very easy to
plan, with minimum width, and only one rail.

The rail was offset 10 ft away from the surface
of the mountain, to avoid digging, so the radius
was exactly 5280 ft. The train could only
ascend a 4% grade. The basestation v/as at
the very bottom of the mountain, so the train
could not get a running start up the hill. How
much track did Little Egbert have to buy? (a)

25.00 miles (b) 25.020 (c) 25.040 (d) 25.40

1 K For extra credit ’ the tra ' n ^aves
A U «the basestation going north on the
west side of the mountain, from which direc¬
tion does it approach the summit? I don’t have
a really solid answer for this one. You tell me!

/>m y Little Egbert was walking across a field
hm4 • at 2.000 mph. The closest road ran
true north, but the place he was going was 2 miles north and 0.1
miles east of his starting point. So he walked straight across the
field, the shortest route between two points. His brother Pythago¬
ras decided to walk along the road at 2.000 mph, for 2 miles, and
then cut true east. When he got to the corner and started east, he
saw that Egbert was quite close to the end point. He decided to
speed up to get there at the same time. How fast did he have to
walk to achieve that ? (a) 8.0 mph (b) 16 (c) 80 (d) 160.1 (e) 442

3 It’s easy to set up two double-pole double-throw switches,

* one at each end of a room, so people entering the room can
turn a common lamp on or off with either switch. Little Egbert
added a new kitchen and dining room to his house. The kitchen
was triangular with three doors. How could he arrange three switch¬
es so that anybody entering at any door could turn the main lights
on and off just by throwing the adjacent switch? The dining room
had four doors. How could he arrange switches at all four dining¬
room doors with the same capability? (Note, I just got a new
kitchen with three doors and a new dining room with four doors.)

4 Here’s another sphere problem. Little
o Egbert sat atop a 16-ft stepladder, in

Quito, Ecuador, astride the equator. He low¬
ered a weight by a thread from a well-defined
point at the top of the ladder to mark the
place on the ground directly below it. Then he
pulled up the weight—and waited—and
dropped it. It hit the ground at a place not
exactly the same as the first point. What was
the difference in position? This goes to show
that when you drop something, it does not
just falll “down.” (Unless you are at the North
Pole or the South Pole.) Assume the radius of
the Earth is 4000.000 miles and that g =

32.0 ft/s 2 . Thus, the rotational velocity of the
Earth is 25,132.74 miles/24 hours, or
1047.20 mph, or 1535.89 ft/s, or 18430.70
in./s. An accuracy of 1% is requested. (If you
have a release mechanism that does not
affect the weight’s transverse motion, you
can find out how far you are from the equator.
Or, you could find out if you have a release mechanism that does
not affect the weight’s transverse motion by rotating the release
mechanism in various directions.) (a) 1.4 mils (b) 14 (c) 34.2 (d) 112
ff* Compute the square root of 156 to a precision better
©than 1 ppm. (It is easy to compute to 1 ppb.)

The answers will be published in Electronic Design. Complete
solutions with explanations will be posted on my Web site at
wvm.national.com/rap on April 15. (I apologize in advance for any
complaints that I should have written this in metric terms. Sorry, but
that’s not going to happen. Brain puzzlers do not start out “There
was a hemispherical mountain with a radius of 1609.265 m...") ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A , National Semiconductor
P.O. Box 58090, Santa Clara , CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara , Calif.

18 ***

03.30.06 ELECTRONIC DESIGN

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

What's All This Bivouac
Stuff, Anyhow?

O kay, my old friends Mai and Jon and I were planning
to hike up a big hill, Mt. Adams, near Gorham, N.H.,
in early February. On snowshoes. We’d have 3000 ft
of snow to ascend, and it gets steep.

We considered the potential terrain,
like a lot of packed snow on the ground,
which would be fine. But would there be
a lot of new snow? A foot of new snow
would really slow us down. Two feet of
new snow would cause some serious
problems, because in a group of three,
each person has to do an average of
one-third of the work breaking trail,
if you have never hiked in deep snow on
snowshoes, you would figure out pretty
quickly that breaking trail is a lot of work.

We all figured this out fast in 1955,
when we started up this same hill.

Even a half-foot of snow is a good, hard
test for a group of six or 10 people, as the
relevant fraction is still challenging. With
only three of us-as well as the possibility
(probability) of one or two feet of new snow-what would we do?
Give up and go home? Heck, no!

I figured out fast, as I contemplated this problem, that we would
have to be prepared to bivouac up on the trail if there were bad
conditions. We would have to set up a minimal shelter, lay out our
sleeping bags, sack out for the night, and restart the next morn¬
ing. We certainly wouldn't want to go back down to our car and
restart from down there. That would be too slow. But I’ve heard of
people who have had to do that, as they had no better plan.

GIMME SHELTER • So, what’s the shelter? A big square of plastic,
about 8 feet square per person, would keep us from getting too
damp or snowy. Wearing our snowshoes, we’d have to stomp down
a shallow, broad hollow into the snow, even if the hillside had a
steep slant. Next, we’d have to put the plastic sheet over our pack-
frames and under our foam pads and then lay out our sleeping
bags on top of that. And, we would have to pull the rest of the plas¬
tic sheet over us-and hunker down for a dozen hours. It’s not
exactly the same as sleeping, but some sleep may be possible.

Also, we would have to plan some sustenance. I’ve invented
some light and quickly heated food-a dehydrated soup with pre¬

cooked sausage and crackers for sup¬
per. On the trail, I would have to fire up
our stove and heat some water. Then,
after some coffee and/or hot choco¬
late plus a couple of breakfast bars for
the morning, it would be back into the
owshoes and up and at ’em.

What if the only good bivouac spot
well below where we want to stop?
We would set up our camp and take a
break. Then we would walk up farther
and keep breaking trail for a while
until we almost got tired, or until it got
kinda dark, and then go back and
sack out. It’s very important to keep
from getting too tired.

After we got to our camp, Crag Cab¬
in, at 4200 ft, we would have no great
amount of trouble with snow. Up above the timberline, the fluffy
snow blows away, mostly. We would just put on our crampons and
walk, per my recent column on that subject (electronic design, Feb.
2, p. 20, ED Online 11919 at www.electronicdesign.com). The
descent would be trivial. Even if we got one or two or three feet of
powder, it would barely slow us down.

So while I have rarely done much bivouacking, I sure know
what to do. All we needed were some good plans and that 7-oz
square of plastic.

How did this work out? We had very little snow, and that was
hard-packed. So we left our tarps in the car and had no trouble.
We never even put our snowshoes on! But we knew we were
prepared for anything. For the complete trip report, see Part 17
at www.national.com/rap/nepal/index.html. ©

Comments invited! rap@galaxy.nsc.com —or:

MaiiStop D2597A , National Semiconductor
P.O. Box 58090 , Santa Ciara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at Nation¬
al Semiconductor Corp., Santa Clara, Calif.

1X0 ...

04 . 13.06 ELECTRONIC DESIGN

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

Bob's Mailbox

D ear Bob: I once had a problem identical to what
Jeff Slefske mentioned (electronic design, Feb. 16,
p. 18). The way he described it, his circuit did not
originally have the power-supply common connect¬
ed to the line cord green wire.

Often supplies have 0.01 \xf or so from each side of the ac
line connected to the common to reduce conducted EMI. If the
common is not grounded it floats to 60 V rms, with enough cur¬
rent available to give a nasty shock and also blow RS232
inputs. At 0.01 jmf, about 0.5 mA is available. (That sounds like
“only 60 Vbut that Is 84 V peak. If you have a 0.01-pF cap
charged up to 80 Vac peak, the momentary surge will be larg¬
er than 0.5 A, which is capable of blowing a lot of computer
inputs—and semiconductors in general. You can make inputs
that will survive such an 80-V surge, but you'd have to add
resistors and diode clamps. And it takes some real engineer¬
ing, some significant expense, and some extra board size to
protect the inputs, /rap) Another good reason to not pull out
the third pin on your power cords!

• Jim Harman (via e-mail)

• Pease; Uh, yeah! Or, you might avoid this by not doing hot-
plug connections, if you don't know what's going to happen.
Thanks for the comments, Jim.

Dear Bob: I have a story you might be interested in about
“Autosave Stuff” (electronic design, March 2, p. 20). Several
years ago, a coworker and I were writing a document (using MS
Word) and had autosave turned on. We were using many of the
features of Word, including drawing and autoshape options for
many complex drawings. Autosave worked nicely, saving every¬
thing every 30 minutes. This went on for two days. At the end of
the second day we tried to load from the autosaved file and
found it corrupted.

(Ouch! And you trusted it to work, so you never checked it
before you trusted it. Presumably, it had worked in the past but
started to go bad when you needed it. If I have an important
document, I will“save to file" and also occasionally “save to
template," which is less likely to get scrambled or erased. I also
print out important pages on paper, where it's easier to proof¬
read with full accuracy—and slightly less likely to burst into
flames, /rap)

We never saw any warnings. And it did not save each new
autosave in a separate file. Ouch! Now when I see that an
autosave is engaged, I manually save to the file name + a
sequence number.

It gives me the feel of being back in 1969 and resaving every
10 lines entered when logged onto timesharing on a main¬
frame. All that time and technical progress, and we still can not
depend on the hardware/software beyond the keyboard and
monitor, be it a mainframe or a PC.

• Charles Ryan (via e-mail)

• Pease: Yeah, I guess you can hardly trust any computers.
Not even analog computers... or punched cards?

Dear Bob: Autosave is one of those things I keep turned off in
MS Word. Consider this scenario: I have just done one of
those stupid things where I somehow delete a block of text
(maybe click on delete instead of cut). The phone rings and I
am on one of those long support calls. So autosave soon
saves the corrupted document. Of course there is a .bak copy,
isn’t there? Sure. But 10 minutes later, while I am still on the
phone, autosave again saves my corrupted document and
saves a backup copy of it as well.

(Ouch! I never had that problem, /rap)

Sure, I could still come back and undo the changes. But I am
less likely to remember to do this than I was to manually save
my work at regular intervals. What is needed is a popup
reminder.

• Frank Delts (via e-mail)

• Pease: Hello, Frank. PC Write Lite used to provide that, too,
on request.

MBVTAL MATH STUFF • Here are the answers to the brainteasers
in my “What's All This Mental Math Stuff, Anyhow” column
(electronic design, March 30, p. 18). For more details, see
www.national.com/rap.

1. 25.0400 miles, or 132,211.087 feet
lb. I don't know, but maybe soon

2. 80 or 80.05 mph

3: Three or four double-pole double-throw switches are fine
(see my Web site for a full explanation)

4:12.48999600

5:14 or 13.97 milli-inches ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor CorpSanta Clara, California.

20 ...

04.27.06 ELECTRONIC DESIGN

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

What's All This License Plate
Stuff, Anyhow?

et’s call this the License Plate Game, because that’s the
easiest place to get three nearly random letters. You
start with three given letters, and you have to make
them into a word. And you start with eight points.

If you add one (or more) vowels to the beginning or
end of the letters, that costs you one point. Or if you
have to add one or more consonants to the beginning
or end of the letters, that also costs you a point. If you
have to add some (one or more) vowels and also some
consonants, your score is down from eight to six
points. It doesn’t make any difference how many let¬
ters you add. But that’s just the start.

If you have to change the order of the given letters
to turn it into a word, that costs you two points. And/or
if you have to insert any new letters into the original
letters, that costs you two more points. So if you have
to do all those things, you are down to just two points.

But if you can’t make any word at all, then it’s not too
surprising that you get zero points.

You can see that this is not at all like sudoku, where every
move is just logical. In this game, you have to be able to manip¬
ulate letters and make words. And as a first hack, I’d say the
scoring is based on elegance.

Now, that isn’t the end of the scoring. Let's say that by using
those rules and procedures, you can make a word that’s worth
six points, or whatever. But you can win one bonus point if you
can take the given letters and make one word using only added
vowels and make one word with only added consonants. And
that is a nice little challenge. I think it's fun. I got the concept of
this game from an old friend, Karen, but she started with six
points. I think eight is fairer.

There is one more side rule. If the three given letters already
make a word, that doesn’t count. You have to change or add
something to make some new word. Now, there may be only
20 3 combinations of letters that will show up on license plates.
But that’s pretty good. 8000 is enough. Note that “ABC” is not
the same as “CBA” or “BCA,” so it’s not too boring.

How can two or more people play in a competitive way? I
haven’t really thought of any firm rules, but obviously you have
to have some plans for allowing enough time for all the players
to find a word or words. Then at a suitable time, somebody has
to decide who states his solution. Maybe the player with the
highest number of claimed points goes last. What words are

tywwjjNSHj Garder

considered legal? Well, you can set your own house rules and
pick any dictionary you choose. Myself, I tend to accept plurals,
but not proper names, such as geographical places.

Here’s an example. RED is already a word, and you can easi¬
ly make REED or REND. However, those involve splitting the
original letters, so those would make just five points. Therefore,
SHRED is better, as it does not involve splitting. And those sev¬
en points, added to the bonus point (for the one word you
made by only adding consonants, and the other word you
made by only adding vowels) brings you back to eight.

When there are awful combinations like JXZ, you concede
quickly and go on to the next car—the next set of letters. At least, I
can’t make a word with J, X, and Z. If you find any combinations of
letters that seem hard, send them to me as a challenge! ©

Disclaimer: This game specifically is not recommended for drivers, who may
get distracted and forget where they need to turn off or when they need to put
on the brakes. Avoid playing this game if you get distracted.

Comments invited! rap@galaxy.nsc.com —or:
Mail Stop D2597A, National Semiconductor
P.O. Box 58090 , Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

20 <’

05.11.06 ELECTRONIC DESIGN

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

Bob's Mailbox

D ear Bob: Just read your recent “Mailbox” column
(electronic design , March 16, p. 20). You comment¬
ed on a reader (Mike Smolin) who stated that his
tire dealer put nitrogen in his tires. There are good
reasons for this. The main advantage is that nitrogen is dry,
whereas compressed air has a lot of water vapor in it. (Ordinary
compressed air may have some water vapor in it A iittle. it may
or may not have much, /rap) When the tires get hot the water
boils and increases the pressure in the tires, which affects han¬
dling, braking, etc.

• Cliff Harris (via e-mail)

• Pease: How often do you run your tires above 150°F? Rac¬
ers may. I don't. I doubt if you do. So do you want to pay $8
or $12 for nitrogen in four tires? Yuppies rnay want to have
"the very besteven if they can't tell the difference. Just like
the Emperor’s New Clothes.

NOTE: Even at 220°F, water does not boil in an
atmosphere with 30 psi of pressure. So how
much “vapor pressure” will they add at
150° F? I’ll guess 3 psi if you have a lot of
water in there (1/4 cup) and perhaps 1 psi if
you had 1/8 teaspoon. For sure, on my VWs,
the tires don’t get above about 150°F, no
matter what. The temp rise is barely 25 °F
above ambient, and even in a desert, it rarely
gets above 125 °F. If they are that hot, your
30-psi pressure has already risen to 34 psi, and
that’s not a big deal (on my Beetle, from 27 psi to 31). If you
are running your tires hotter than that, well , you probably
have other problems. You said there were “good reasons.’’
Good reasons for the mechanic? Like what—boat pay¬
ments? I’ll just keep air in there. For racers, balanced infla¬
tion is very critical, so they use dry nitrogen to prevent any
3-psi imbalances.

Dear Bob: My thoughts are that the new tires always go on the
front. I put a very high premium on being able to steer. (Yeah,
but these guys argue that being able to steer only a little, and
to skid not at all, is better. Better for them, maybe, /rap) Only
a real yahoo would put the new ones on the rear of a front-
wheel-drive car. (That is exactly what these guys think is
safest, /rap)

I miss rear-wheel-drive cars because they always wore the
rear tires faster. Then you would put the new ones up front and
have nice new ones for steering and half-worn ones on the rear
for decent traction. I like a little oversteer! Always hated front-
wheel-drive (a Saab, two Toyota Camrys) because of its some¬
times abrupt understeer. (Yeah, but that is about all you can
rent these days./rap) Remember that stupid myth about front-
wheel-drive: “If you go into a skid, floor it—it T ll pull you out”?

Talk about crapola.... (Check. Have you seen my pretty good

BOB PE!ASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

book, How To Drive Into Accidents...? Check it out at
www.transtronix.com. I put at least half of the chapter on skid¬
ding on the Web. Comments invited, /rap)

I’ve been driving Subarus for years, now, and you can’t beat
four-wheel-drive. Plus the torque is biased towards the rear, so
you can kick the rear around if you want to.

• Nick Allen (via e-mail)

* Pease: Hello, Nick. Well, you and I agree that we want good
traction under our drive wheels. But a lot of other people
wrote in to say that it is dangerous to have worn tires on the

rear, even for
front-wheel-drive
cars. They argue that
most drivers are used to understeer and plowing. But under
braking or under cornering with slack throttle (especially on
rainy or slick surfaces), this can change to oversteer and a
nasty skid. Maybe most drivers would be safer if they avoid¬
ed that. Maybe they're terrified of getting into skids and
don’t know how to get out of skids.

The chances that I will own a front-engine car or front-wheel
drive are negligible. But that's what some worrywarts worry
about, and apparently the tire companies do, too. So if you
take in your car, and they insist on putting the new tires on
the back, you will have to overrule them by rotating your
own tires, front to rear, right in their parking lot and thumb¬
ing your nose at themI (And then you have to be extra care¬
ful, for a while, not to skid into any accidents.) ©

Comments invited! rap@galaxy.nsc.com —or:
Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

20 • •

05 . 25.06 ELECTRONIC DESIGN

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

WHAT'S ALL THIS "ERROR
BUDGET" STUFF, ANYHOW?

I was just on the phone explaining how to do an “error
budget” analysis on some fairly simple circuits to a young
engineer. Later, I mentioned this while I was visiting with
my friend Martin, and he said he had been quite surprised

El * ®#t +°

ZLt. _I-ap

All K'S ±!%

f/jur* 2 .

when he found that many engineers in Europe were
quite unfamiliar with the concept of an “error budg¬
et.” How can you design a good circuit without being
aware of which components will hurt your accuracy?

When I was a kid engineer back in 1962, my boss
George Philbrick gave me a book on differential
amplifiers by Dr. R. David Middlebrook, and he
asked me to do a book review. I studied the book,
and it was full of hundreds of partial differential
equations. If you wanted the output of a circuit with
14 components, you could see a complete analysis
of how each component would affect the output off¬
set and gain. Each equation filled up a whole page. It
did this several times.

Yet it didn’t offer any insights into what’s impor¬
tant. I mean, is (3 X d(Rl) more important than R1 X
d((3)? In retrospect, I’m glad I didn’t submit any cri¬
tique of that book. I woulda done more harm than
good. Such a mess! Even now, it would be hard to
write a critique on a book that was so true, but so unhelpful.

Things are much simpler now that people are mostly (but not
entirely) designing with op amps. The best thing is that the out¬
put offset and dc gain and ac gain errors are largely orthogo¬
nal. An “operational” amplifier does perform, largely, an “oper¬
ation” based on what task you ask it to perform when you
“program it” with Rs and Cs. If the offset varies, the gain does
not, and vice versa. We all agree that it's very helpful that you
can compute what the performance will be with almost no
interaction. No partial derivatives.

Now, let’s take a look at a couple of applications—real cir¬
cuits—and their tolerances within an error budget. Here is an
amplifier to magnify the I x R drop of current through a O.l-Q.
resistor and bring it back down to ground. Figure 1 shows a
conventional differential amplifier, with the common mode up
at+12 V. The gain of-20 will bring the 1.0 A x 0.1 Q. signal
down to a ground level. If the current is 0.1 A, the output will be
0.2 V, “small-scale.” A full-scale current of 1A will bring the out¬
put up to 2.0 V, which is suitable to send to a detector or ana-
log-to-digital converter.

Let’s select an op amp like the LMC6482B, with low offset
voltage less than 1.0 mV. (There are other versions of this
amplifier with less than 0.35 mV, but let’s select an intermedi¬
ate model.) This 1 mV does cause 21 mV of output error. This
op amp has less than 20 pA of l B at all temperatures, so at
least that’s negligible. (Bipolar op amps might have small l B
errors, but you’d have to check it.)

Now let’s see what the resistors add. Assuming all Rs have a
1% tolerance, the gain of (2.0 V per A) has a tolerance of ±3%.
This would cause ±60 mV at full scale, but only ±6 mV at "small
scale” (0.1 A). This may be acceptable.

Then let’s consider the common-mode errors. If R4 has a 1%
tolerance, and it has 11.4 V across it, the 1% tolerance could
cause a 114-mV error. By symmetry, a 1% error of each of Rl,
R2, R3 can cause another 114 mV! Added together, the com¬
mon mode could cause an output error of 456 mV! That’s
about ±1/4 of full scale—even for small signals. That doesn’t
look so good to me!

It's true that if adjacent 1-kft resistors are inserted, they’re like¬
ly to match within ±1/2% so the probable error between the pair

18 •••

06 . 08.06 ELECTRONIC DESIGN

Pease PORRIDGE

might cause ±60 mV, and the ±1/2%
matching between the 20 kQ.s would
cause another 60 mV. That added to the
21 mV from the V 0 s would add to 141 mV.

Some textbooks teach you that you
should add these errors arithmetically to
141 mV. Others point out that they could

20 ...

be added in an RMS way, so that 60 + 60
+ 21 mV = 87 mV. Typically, this might be
true. But the worst case of 141 or 456
mV might be more realistic. I mean, if
you’re going to build 1000 circuits, and
most of them are better than 141 mV,
what are you going to do with the 400

circuits that are worse than 141 mV?
And, that’s still 7% of full-scale....

You could go shopping for 0.1% resis¬
tors, but they aren’t cheap. You could put
in a trim-pot to trim the error to (no offset
error) for small signals. But as you may
have noticed, a trim-pot has to be proper¬
ly trimmed. And if that pot is accessible,
it could someday be mistrimmed, and it
would have to be corrected, in some awk¬
ward calibration cycle. Most people want
to avoid that trim-pot. Before we decide
that this 141 mV is unacceptable, let’s
look at another circuit.

Figure 2 shows an alternative circuit
with the same gain, 2.0 V per A, using a
PN4250 or 2N4250, a high-beta pnp
transistor. What does the error budget
look like? The same op amp causes just
20 mV of output error. The 1% resistor
tolerances cause the same gain error,

60 mV at full scale, or 6 mV at “small
scale.” The newly added transistor adds
(-1/3%) max from its alpha, or less than
7 mV, at full scale.

What is the offset error due to com¬
mon-mode rejection ratio (CMRR), or due
to resistor mismatch? Nothing. Zero. The
transistor doesn’t care about the voltage
across it. There are no resistors with 12
V across them.

So the offset error is ±20 mV, due pri¬
marily to the amplifier's V 0 s (which could
be reduced), not ±400 mV. This little cir¬
cuit has greatly reduced errors com¬
pared to Figure 1, even if Figure 1 had a
couple bucks of 0.1% resistors. This may
be acceptable. Even the offset errors
could be reduced to 7 mV by selecting
the LMC6482A.

So we have seen that circuits with simi¬
lar functions can have completely different
“error budgets.” I love to recommend
amplifiers with high CMRR. But depending
on cheap 1% resistors can hurt your “error
budget” a lot more than you’d suspect. ©

Comments invited! rap@gafaxy.nsc.com
-or:

Mail Stop D2597A
National Semiconductor
P.O. Box 58090
Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961
and is staff scientist at National Semiconductor
CorpSanta Clara, Calif.

06.08.06 ELECTRONIC DESIGN

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

Bob's Mailbox

D ear Bob: When looking at datasheets or appli¬
cation notes that have the schematics of an
1C in them, I often see very strange compo¬
nents; transistors with two, three, or even
four collectors or emitters, and a component that looks
like a capacitor where one plate is a resistor. The LM875
datasheet on page 5 has both of these. For example,
check out Q12, Q15, Q16, and R23. Application Note
446B on the LM12 internal design also has these. On
page 3, Figure 3, Q3 and Q4 have two collectors and a
short line through an extension of the base that seems
to be shorted to itself. And, R32 in the lower right has
one of these resistor-capacitor hybrids but nothing con¬
nected to the plate. I’d like to learn more about what
these mean and was wondering if you could write a col¬
umn about them and/or point me at some references
that talk about them.

• ten Fischer (via e-mail)

• Pease: Hello, Len. Okay, I finally found a few minutes to
peek at the LM675 datasheet. It Is perfectly logical. The
pnps that seem to “have four collectors” really do have
four collectors. Whatever current Is sent Into the emitter
winds up at the four collectors, split into about four equal
parts. I’ve been doing that for many years. My LM331 has
done this since 1977, and the LM301 and LM741 have
done this since the early ’70s. So has Tom Frederiksen’s
LM324. If you wanted to see more about this, you could
look at www.designarrays.com , which is Hans Camen-
zlnd’s Web site. The resistors with the bar above them are
pinch resistors. You can iearn about them also at Hans’
site. They have lousy tolerances and lousy tempcos, but
they’re better than nothing. This is nothing new. You’re
just starting to notice them, but they have been around
over 30 years. And, I finally got a chance to look at the
LM12 datasheet. The resistors with bars are pinch resis¬
tors. The bar across the collector of a transistor repre¬
sents a ring that collects current when the transistor sat¬
urates. I had to do some digging around to find the
answer to that one.

Dear Bob: Have i had fun. I recently took up a course on
arc welding (one needs to stay entertained), and wow! This is
something I think every electronics engineer should experi¬
ence. Just you, a 25-V, 150-A arc, a mask, a stick, and a littie
voice in the back of your head going "PULL THE POWER!
PULL THE POWER! IT’S BAD! VERY BAD!”

• Dan Williams (via e-mail)

• Pease: Yup, there sure are a lot of of femtoamperes in a
150 A arc! “Billions and billions!” I welded up a tandem
bike, starting from two bicycles, when I was a kid. I
rode it several hundred miles, and it was stolen just

about the time the main welds began to go. Yeah, weld¬
ing is a useful art—like baking or cooking.

Dear Bob: (Forwarded to Bob by Ted Hardin, senior appli¬
cations engineer at NSC) Could Bob Pease comment on "the
restriction of the use of certain hazardous substances in
electrical and electronic equipment (RoHS)”? Are there that
many old lead-filled scopes going into landfills? The new no
lead solder really sucks. My technicians having to use this
stuff are cursing me out. It is too stiff, doesn’t wick well into
thru holes, and always looks like a cold solder joint. Are there
any good choices for nolead solders?

• Dan Mohr (via e-mail)

• Pease: Hello, Dan. I agree, it is unfortunate that we
have to use inferior solder. I’m not sure exactly
where it says you can’t use solder containing lead for
experiments. But exactly where that crosses the line
to “not using it for production,” or for material that Is
to be shipped out, is something I have not heard. I’li
look into that. But why ask me? Ask Dear Abby. Ask
Kester or other people that make solder. I don’t sell
any solder. Let me know if you find anything good.
Meanwhile, get a hotter soldering iron!

Dear Bob: I just finished reading "What’s Aii This Bivouac
Stuff, Anyhow?” (electronic design, April 13, p. 20). So, what’s
all this snowshoe stuff got to do with Electronic Design, any¬
way? (if I froze to death, I couldn't write any more columns.
That's what./rap) In short, planning and preparation. That’s
the key to success in any venture, and the lack thereof can
be the key to failure. Works the same for manufacturing elec¬
tronic products or climbing mountains. Thanks, Bob!

• George Gerwlck (via e-mail)

• Pease: Check! /rap ©

Comments invited! rap@galaxy.nsc.com -or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara , Calif.

06 . 22.06 ELECTRONIC RESIGN

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • • rap@galaxy.nsc.com

What's All This Theft
Stuff, Anyhow?

y wife got stung recently. Somebody stole her com¬
puter’s battery charger from her checked bags at
San Francisco International Airport. Then I got
stung the same way.

As I was leaving my hotel in Phoenix, Ariz., I tucked my
computer’s battery charger into the outside pocket of my
unlocked suitcase and left to go to Sky Harbor International
Airport. I didn’t put the charger in my knapsack to carry it with
me onto the plane as I usually do because the computer had
a good charge and I didn’t expect to do much typing. After I
arrived in Los Angeles and went to my hotel, I looked for my
charger. It wasn’t there.

So, I was able to borrow some charge from my buddies’
chargers and do just a little typing for the next couple of days. I
survived. When it happened to my wife, she was annoyed. Ever
since it happened to me as well, I’ve been very annoyed.

The pattern is obvious. The people who run the X-ray
machines at the airport are passing on the word—by some
kind of sticker or chalkmark—to some accomplices who know
where to reach inside the unlocked suitcases and grab a
$100 item. Most travellers won’t even report this to police,
but I will. So far, several colleagues have said that battery
chargers have been stolen from their checked baggage, too.
So, I’m not just imagining things.

If you’re carrying a laptop and you have a battery charger, or
anything of value that is metallic, don’t put it in your unlocked
suitcase. Carry it with you. Yes, I know a pound is a heavy load
to add on to your carry-on baggage. But the avoided inconven¬
ience is worth it.

If your battery charger is stolen, report it to the police. Eventu¬
ally, we will nail down the culprits. They’re hurting us $100 at a
time, but we will apprehend them. Or scare them into stopping. If
you check in your baggage at an airport, the guys who run the X-
rays know what’s in it. Some of the guys at the arrival airport
might like to steal what’s in your baggage, but they don’t have X-
ray machines. So you can tell who has the info.

I could buy an FAA-approved padlock and put it on my suit¬
case. But that’s a joke with a soft-sided suitcase. If I put a $20
lock on a $15 suitcase, the thieves would be all the more curi¬
ous about what I had. They’d just slash through the side. Nor¬
mally I don’t keep anything of great value in my suitcase.

Specifically, I always carry my lecture notes in my brief¬
case. I would never put them in my checked baggage. Even if

they’re of no value to
anybody else,
they’re very valu¬
able to me, as I
couldn’t do my
lectures with¬
out them. Even
if they were
delayed, I’d
have prob¬
lems. Except
for a clean
shirt, there’s
nothing in my
suitcases that I
can’t do without for a
day. I could carry a
clean shirt in my
briefcase, but I
don’t.

I have a hard¬
shell suitcase, but there’s no place to add an external lock.
Security? At an airport? No further comment. I read recently
that 87 thieves have been caught at airports. What about the
other 870 (out of the 87,000 honest airport workers) who are
still unapprehended? They may be honest most of the time, but
a few of them are surely ready to grab a valuable item. My suit¬
cases met up with one of them.

In general, I don’t leave valuable things in my car. I know how
easy it is for a guy with a coat hanger to get into a car. If I’m
going to see the dentist on my way home, I bring my briefcase
with my laptop into the dentist’s office.

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

ED ONLINE 12940

20 ••

07.06.06 ELECTRONIC DESIGN

• •

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

Bob's Mailbox

This motorcycle road-racing engineer
can tell you that the whole nitrogen-fill thing is
usually a load of you know what (see “Bob's
Mailbox ," March 16, p. 20). The first thing they
do when demounting/mounting your tire is to spray the
entire bead area with copious amounts of soapy water.
Repeatedly. So all the dry nitrogen in the world isn’t going
to amount to a hill of beans when the environment inside
the tire is quite moist. To make it worth anything you’d
need to use a commercial non-water-based tire mounting
lubricant. So for those who need the very best, they should
make sure that they get the right lube with that nitrogen!

• Scott Traurig (via e-mail)

• Pease: Hey, Scott, I forgot about that!! Maybe if they ran dry
nitrogen through the tire (in and out?) after it was mounted,
they could eventually dry it out. Inflate/deflate, nine times?
Absurd!!! Maybe if they used some solvent that was not
water-based, they could avoid the moisture. Maybe racers
do that. I mean, I understand why they do want dry air—to
minimize changes in inflation. Would they use some kind of
rubber cement? Next time I buy a tire, I'll stand there and
watch them mount it. Then if they wanted to sell me dry
nitrogen, I'd ask why they wet it down first. But nobody has
ever asked me if I wanted nitrogen. Not yet. Thanks for
being so observant. I don't buy tires very often or have them
remounted. I usually bring in a flat and tell them I'll be back
tomorrow fora new tire....

Hiya Rob/Bob: I’m currently doing an undergraduate thesis
on reference circuits at the University of New South Wales, Aus¬
tralia. My search on bandgap references brought me to your
page and the page about Bob Widlar. I just wanted you to know
that it is rather refreshing to see engineers with personality
and flair, and hopefully I will be able to enjoy what I do when I
finish my degree as much as you guys. All the best.

• Min Sun (via e-mail)

• Pease: Hello, Min. When I came to NSC, slightly over 30
years ago, Bob Widlar was just preparing to bring out the
LM10. That was one hell of an op amp with 85 transistors. I
am still lecturing about the latest and greatest version of
the LM10 op amp this week (as I write this) and next week.
And about its reference. When you graduate, if you think
you got “personality and flair" and circuit talent, the analog
engineering world needs you!

Dear Bob: Just read your piece on dielectric absorption (DA)
at www.national.com/rap/Application/0,1570,28, OO.html.
Very useful information there. However, I still have a question.
How does one know the DA of conventional surface-mount
caps? (It’s easy. Charge it up to +10 V, for perhaps a second.
Then short it out fora few milliseconds with a switch. Watch its

Vqut rise up, using a high-impedance voltmeter, or a good op
amp with l B less than 1 pA. If it rises a few mV, it's NPO. If it ris¬
es dozens of mV, that's some inferior ceramic, /rap)

Our application uses an LF398M sample and hold, with a

O. 01-jiF cap for an acquisition time of about 20 jis and hold
step of 1 mV. But there are no curves in the Typical Perfor¬
mance Characteristics graph for surface-mount caps. Any idea
on that, or is it time to build up a test rig?

(The test rig is very simple. Set up the LF398 with +10-V input,
and sample fora second or two. Then go to hold. Then set the
V iN to zero, and put in a short sample pulse, perhaps a few mil¬
liseconds. Watch its V 0UT . There will be a big difference. The
LF398's leakage of perhaps 50 pA will cause a leak rate of 5 mV
per second. That is good enough. Note: If you do not charge the
X7R cap up to 10 V, then the lurch after it is shorted will not be
bad, and it would be hard to tell an X7R from a COG. /rap)

The input signal is coming directly off an AD622 instrumen¬
tation amp connected to a photodetector. And the signal level
is about 10 mV. (The LF398 would do a better job if there were
a preamp ahead of it to bring that 10 mV up to 100 or 200 or
500 mV. /rap)

• Dexter Francis (via e-mail)

• Pease: Hello, Dexter. Most 0.01-pF ceramics are not NPOs,
and they have lousy DA (soakage). Ask your purchasing peo¬
ple what they bought/are buying—most likely, X7R types. NPO
ceramics that are 0.01 pF are found in 1206 packages (3.2
mm by 1.6 mm by 1.2 mm tall). You can look it up in Digikey—
and they cost $0.40 to $0.90 in quantities of 100! Most
0.01s are in a size 0204 or 0306 and cost a dime or less.
They would be X7R or Y5V. And they have lousy DA. ffi

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor

P. O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

18 ••

07 . 20.06 ELECTRONIC DESIGN

• •

Pease PORRIDGE

>E E, Contributing Editor

• • • rap@galaxy.nsc.com

What's All This
Penny Stuff, Anyhow?

friend wrote to me to complain
about a new proposal to lock out
individual Internet users. It seems like
the issue is costs versus access. Big
companies would be able to send and receive
e-mails faster and cheaper than individuals.

If this plan comes closer to reality, I will complain more
vigorously.

“I don’t usually write to congressmen, but I do when it’s
important. Right now, I’m not sure this is important. Not yet,” I
replied to my friend.

“I would not mind paying a penny for every e-mail I send. I
would even pay, reimburse, my company, NSC, a penny for
every personal message I send,” I continued. “In the last two
months, I have sent over 1000 e-mails, and let’s say half are
personal. Would I change my e-mail habits if I had to pay $2.50
per month? Heck, no. I spend more than that on stamps.”

I wouldn’t even mind paying a penny for every search I do. I
mean, every search I make takes away several seconds of my
life as I wait for the results. So I don’t search for things lightly.

By the way, I did find some things in a recent search that
weren’t there a year ago. I usually use dogpile.com, but I must
admit, google.com did find the lyrics to Shango’s “Day After
Day (It’s Slippin’ Away)” when Dogpile did not. “What can we do
with a bushel of wet gold?” the group sang in 1969, when the
song hit #57 on the national charts.

I would be delighted if all the spammers had to pay a penny
per message. That might cut them back a lot. I’d vote for that. I
won’t approve of any volume discounts for them. And, a penny
per megabyte per addressee would seem fair. Why would we
protect the pigs who clutter up the airwaves? (Or, okay, wires.)

You’re familiar with the idea that some people wouldn’t
stoop to pick up a penny. But picking up pennies, one per sec¬
ond, even if I had to stoop separately for each one, could pay
me $36 per hour—and that’s not bad wages. Furthermore,
these are pennies with no tax, so each one would be worth 1.7
cents or so. So, pennies aren’t trivial.

Nor is my time trivial. I could save a few pennies per day if I
drive slower, but that would waste a few minutes a day. I am
holding at 65 mph as a reasonable compromise. I used to drive

at 68 mph, and I will again if gas gets below $2.99. But I am
jealous of my time.

As for computing access, I think this proposal is dumb.
Somebody will squash it, sooner or later. I already have high¬
speed access when I’m at work. My company pays for it. When
I’m at home, I have a mediocre slow modem, but that’s okay
with me. Most people who aren’t willing to pay for a lot of fast
access get slow, cheap access. That’s okay with me too. I can’t
see a big deal about what you pay and what you can get. I don’t
think that’s a real problem.

TYPING TRICKS ¥ I just used my left thumb on the space bar—
which is contrary to good typing practice. But when your right
hand is doing several other tasks, backspace and end, it may
be okay.

Did you see in my column that many computer guys make @
by hitting the shift key with their left little finger and reaching
over to the 2 key with their right index finger (electronic design,
Aug. 4, 2005, p. 20)? I must say, I do. I can do it the other way,
but it is a forced deal.

Hey, do you know what’s funny? My stupid computer has
locked up (for 45 seconds) about once every 20 minutes for
the last week. And I just noticed that it has stopped doing that.
That sure is fine with me. I have tried a dozen tricks. I also have
asked several computer experts to help. They tried, but it did
no good. Now, it seems to have gotten better at a random time,
and for no good reason. (Merde.) ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090 , Santa Clara , CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara , Calif.

ED ONLINE 12978

18 ••

08.03.06 ELECTRONIC DESIGN

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

What's All This C-R Stuff, Anyhow?

resistor is a resistor, and a capacitor is a capacitor,
right? Maybe, and maybe not. I’m doing some research
in audio circuits. I’ve heard all the scientific claims that
if two circuits measure the same, they ought to sound

the same. This claim is refuted by the observa¬
tion that they don’t sound the same. That’s a
pretty convincing rebuttal. I gotta believe it.

I’m an analog and measurement guy. If
somebody says some audio circuits don’t
sound the same, I bet I can measure the dif¬
ference. I thought of the old claim (by people
with “good ears”) that electrolytic capacitors
in C-R coupling networks don’t sound the
same as high-quality film capacitors in an
ordinary audio circuit.

I could set up some C-R coupling circuits
and measure the differences in the outputs, if
any (Fig. la). The capacitors might be poly¬
styrene, mylar, or electrolytics (1 pF and 100 kQ). I could use
some precision differential amplifiers to see the “error volt¬
ages”—the difference between the inputs and the outputs. Yeah,

I could do that.

But I’m a lazy guy. I can see that same “error voltage” if I mere¬
ly swap the R and the C and look at the voltage across the capac¬
itor (Fig. lb, basic low-pass filter). It’s a matter of viewpoint—
what point you define as ground. I could watch those small error
voltages on an ordinary scope and compare them, and I could
even subtract them. So I set that up.

This filter has an f (3 dB) of 1.6 Hz. Shouldn’t that be far
enough away from 20 Hz at the low end of audio frequencies? I
put in 120-Hz sinewaves, triangles, and square waves. It was kin-
da boring. I couldn’t see any difference. Then I
cranked the square wave down to 12 Hz.

The two waveforms were different. They
matched for the first 20 ms, and then the elec¬
trolytic had more curvature (Fig. 2, lower trace)
as if its early capacitance was 1 pF but later
changed to 30% bigger. I’ve used a lot of
capacitors in my day, but I never expected the
capacitance to change oddly with frequency
like this. Would this change if I swapped scope
channels? Nope.

What if I changed the R or C values? Well, I’m
an old analog computer guy, and I didn’t have to
put in more capacitance. I just changed the virtu¬
al frequency back to 120 Hz. The capacitors may

not be perfect, but when the errors are very small, can anybody
hear them? And will anybody care?

That may lead to a test where even guys with good ears can’t
hear the electrolytics as they claim they can. I’ll build up some
circuits with National’s new ultra-linear LM4562 audio opera¬
tional amplifiers, which have less than 0.00003% distortion at
1 kHz. (For the LM4562, go to www.national.com/rap and
search for LM4562 after Sept. 8.) I’ll set up some A-B compar¬
isons and measurements.

Can we hear the difference between mylar and poly? Can any¬
body hear the difference? I can’t. But maybe I can show that in a
good circuit, nobody can hear the difference between mylar and
poly. Or paper. What about 1-pF ceramic? Comments later.

If you’re an audio enthusiast, you may be
interested in attending my Master Class at the
Audio Engineering Society Convention in San
Francisco, October 5-8. For more details, see
www.aes.org. While I’m there, I’ll present a lot of
audio experiments that you can use to evaluate
circuits and components. Y’all come! ©

Comments invited! rap@galaxy.nsc.com —or:
Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is
staff scientist at National Semiconductor Corp ., Santa
Clara, Calif.

ED ONLINE 13166

18 ••

09 . 01.06 ELECTRONIC DESIGN

ED ONLINE 13357

Peas

BOB PE

D orridge

CONTRIBUTING EDITOR

rap@galaxy.nsc.com

Bob's Mailbox

i Bob: In today’s Wall Street Journal, there’s a full arti¬
cle on A4 concerning hydraulic hybrids. It was devel¬
oped by the EPA, which has a small lab. UPS trucks
are being used fortesting, but garbage trucks will be
the first commercial use. (I wonder what the standard“driving
cycle” is fora UPS truck. It must be quite different from the EPA
cycle for cars. Many parts of that cycle might have a lot ofstop-
and-go, but not all—not to mention the typical garbage truck cycle,
which must be pretty wild. Gallons of garbage per gallon? But
around here, garbage trucks already use hydraulics. So for their
pickup cycle, they only need a small engine and a small accumu¬
lator. To haul to the dump, they need more power on the road,
/rap) Ford had been a partner but decided to go electric hybrid
(with its Escape). This is good, but I feel that ultimately, electric is
the most reliable, most efficient, and most elegant solution.
Motors and controllers continue to improve (Oh come on! Electric
motors and controllers have been very good fora dozen years.
Even 100 years ago , they weren't bad. It's just the batteries that
are overpriced or undercapable , regarding range and life./rap)
and even batteries (although I wonder about lithium). (Everybody
says Sony's superior Japanese engineering and Chinese manu¬
facturing expertise are always the best. I guess Dell found out dif¬
ferent... /rap) Electronic Design had an interesting Techview on
the Tesla Motors eCar (see www.electronicdesign.com, ED Online
13201). Now that looks like the future to me!

• Dennis J. Eichenberg

• Pease: You just tell me when the price of those lithiums
comes down by an order of magnitude. I think that may
happen quicker than fuel cells coming down two orders of
magnitude in price , but I'm not holding my breath. Some
idiots were prattling along about the upcoming fuel cells
replacing batteries in laptops. What a bunch of poppy¬
cock! (And you do know the etymology of that word!) Okay ,

I am in favor of many ways to conserve energy in vehicle
operation. Electric cars can do it. Hybrids with batteries
and even (in concept) flywheels can do it. Fine. Hydrogen
fuel cells as a hybrid with batteries? Great , if we can
afford them. Supercapacitors? Well , fine , if they work. Now
this morning on NPR , / heard about a new demo truck
from UPS. It uses hydraulics and compressed nitrogen for
short-term energy storage. The preliminary claims are a
70% savings on energy. Presumably this is for some urban
cycle. Well , that's fine by me! Anything that works!/rap

Dear Bob: We’ve been soldering by hand for well over 50 years
between my technician friend and I. (I've been soldering for
over 50 years all by myself, /rap) We have tried several “new
solder formulations that comply with RoHS and WEEE direc¬
tives. The best replacement, from specifications and perform¬

ance, is the IA-423 formulation (Sn/Ag4.7/Cul.7), which is
eutectic like Sn/Pb37, and makes a nice clean-lookingjoint
provided sufficient flux is used. This solder requires a nominal
(10°C) increase in tip temperature or increased tip-contact
time. Also working well is Kester’s Sn/Ag3/Cu0.5 formulation.
Although non-eutectic, this solder “wets” later than its leaded
predecessors and flows through holes better than any leaded
solder we’ve used. Final finish is not nearly as “cold looking” as
some others. Both of these formulations work very well for hot¬
air rework and assembly of surface-mount devices, which is
even more critical than through-hole these days. It is worth not¬
ing that the IA-423 formulation falls under patent 5,527,628
(July 1993, USA Only), which covers the joints produced
(SN/Ag3.5-7.7/Cul.0-4.0), and the Kester formulation is
patented (JP 50 50 289, March 1993). Non-patented alloys,
such as Stannol TO (Sn/Cul.O), produce the confusing, inferior¬
looking “cold” final finish that was referred to in your June 22
column (see ED Online 12630). Perhaps there is some patent
paranoia that is holding the lead-free world away from these
quite usable alloys. (I will check them out. Thanks for the
advice./rap) I build my own prototypes to this day, even as a
department head. I find soldering quite therapeutic.

• Martin Mayer

• Pease: I've been doing a lot of solder recently and it still
feels nice, /rap ffi

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090 , Santa Clara , CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara , California.

20 ••

09 . 14.06 ELECTRONIC DESIGN

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • • rap@galaxy.nsc.com

What's Al This PLL Stun, Anyhow?

can’t believe it took me so long to write this column. I
designed some good, low-frequency phase-locked loops
(PLLs) about 33 years ago and wrote them up but never got
it published. When I moved from Philbrick to National 30

years ago, I got some good voltage-to-frequency converters
going, using the new LM331.1 also wrote application note AN-
210 (still in print) about 1978, which you can see at www.nation-
al.com/an/AN/AN-210.pdf.

I had a 10-kHz voltage-to-frequency converter. Testing it was a
slow process, as the test technician had to trim the frequency to
20.0 Hz at 20.0 mV of V, N . He also also had to trim to 10,000.0
Hz at 10,000.0 mV, using coarse, medium, and fine trim pots. By
multi plying the frequency by a factor of 10 to 200 Hz, I got a reso¬
lution better than 0.01% of full scale in much less than Is. So, it
made a much better throughput with faster testing.

Later, a guy asked me for some help on a “low-frequency” PLL. I
asked how low. He said 20 MHz. I explained that as far as I was
concerned, 20 Hz was a low frequency. I can do PLLs much lower
than 20 Hz. I can do 200 or 20 mHz, and if I was challenged, I
could probably do a PLL at 20 pHz. I later made a 200-MHz PLL,
controlled by a 5-MHz crystal clock. It was... challenging.

I was reminded of this when my wife asked me to take a cook¬
ie sheet of cherries down the cellar stairs to the freezer. I told
her, “And I have to be careful to not flip them all over.” When I’m
walking up and down the stairs in the dark, I recall that I’m walk¬
ing like a PLL. I lock my speed, frequency, and phase to that of
the stairs going by. I don’t need to see what the stairs are doing
every millisecond. I can feel where I am.

When I’m running up and down the stairs, I rely on my PLL
skills to put my feet in the right places (see “What’s All This Con¬
ditioning Stuff , Anyhow?” at www.electronicdesign.com, ED
Online 9726). Of course, running up and down stairs is only good
for your health if you don’t trip, fall, and crash.

Here at National, somebody put up placards saying, “Use the
handrails to walk on the stairs safely.” Whenever I see that, I
immediately take my hand off the handrails and continue to con¬
centrate on walking down the stairs carefully.

When we went hiking in Nepal to the Annapurna Sanctuary in
October 2005, we knew there would be a lot of stone steps. I
counted them carefully. On the last two days, we descended
8820 stone steps from Ghorapaani to Birethanti. These steps
are known as Gurung Staircases, and famously so. See the com¬
plete story at www.national.com/rap/nepal/annapuma.html.

Over the 14 days of the trek, we ascended and descended
35,510 stone steps. In the Annapurna area, the step sizes are
fairly consistent, so you don’t have to worry so much about the

Bob, Vic Scheinman, and Jeff Nilles stand 12,000 feet below the summit of
Mt. Annapurna. This place, at 14,000 feet, is about 17,500 steps up from
where the trekkers left the cars. pdobyie fNi^

step size. But you still have to pay some attention to where you
place your feet. However, don’t expect consistent step sizes in a
house or hotel in Nepal. They often change wildly!

When we went on linear seminar tours about four years ago,
we used to tell audiences to avoid using PLLs when they wanted
a clock with low jitter. Now we tell designers to use a good PLL
and engineer the loop filters carefully for a clock with low jitter.

NSC has a good Web site on PLLs at http://webench.national,
com/appinfo/wireless/webench/index.cgi. It can help you engi¬
neer your loops. Some of the advice offered for 2400 MHz also
may be applicable at 2400 mHz. I haven’t exactly told you much
about how to engineer good PLLs, but I may have pointed you in
the right direction. These Web sites can be helpful. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

ED ONLINE 13785

18 • •

09.28.06 ELECTRONIC DESIGN

• •

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • • rap@galaxy.nsc.com

What's All This Sallen-Key Stuff, Anyhow?

ecently, people in various publications have been
pointing out that using an ordinary op amp in a
Sallen-Key filter can cause problems. A typical
op-amp circuit, as shown in Figure 1, can have a

frequency response that rolls off nicely above 1 kHz, at 12 dB per
octave, down to -40 or -50 dB. Then the response may roll back

If you choose a fast
op amp with high l b ,
you might choose low
values for Rland R2
to minimize error due
to l b X R. This might
lead you to use large
capacitors, which are
more expensive.

The real problem is
that at high frequencies, the input signal couples through R1 and
Cl and forces current (potentially, several mA) into the op amp’s
output. Real op amps usually have very low Z 0U t at dc. But at high
frequencies, their ac Z 0 ut isn’t characterized. Yet if you ask any op
amp to put out current at higher frequencies, and the gain keeps
rolling off at 6 dB per octave, that indicates that the Z 0U t is going
to roll up with frequency like an inductor.

It’s not hopeless. There are several things you can do.

First, select an op amp with higher input impedance and lower
l b , maybe a CMOS op amp with l b less than 1 pA. This will let you
pick higher values for R1 and R2. In turn, this lets you use smaller
(cheaper and/or higher-quality) capacitors for Cl and C2.

So if R1 was a low value (lk) and fed a lot of current into the
output at high frequencies, first, increase the R values by a fac¬
tor of 10. This improves things tenfold. If you have a 1-kHz rolloff
frequency, for example, the capacitors could decrease from 160
nFto 16.

Now, do it again! Goto 100 kQand 1.6 nF. The theory of the fil¬
ter works just fine when each resistor is increased by a factor of
n, and the caps are likewise shrunken by n —not rocket science.
Now the rolloff goes down a lot further, and the capacitors are
smaller and cheaper.

up or stay flat at higher frequencies.

a JZ

o.i^T

NEXT TRICK ¥ Do n’t just take R1 as a fixed entity: Break R1 into
pieces. If you had 159k, you could break it into 9k (5%) and
150k (1%). Then from that juncture, throw in C3 = 0.01 or
0.0047 pf to ground. This now makes a three-pole rolloff. That
will further roll off the frequency response of the filter, and it will
further decrease the amount of current fed into the output via
R1 and Cl. This added R-C will add even further to the rolloff
and high-frequency attenuation, if you want a lot of that.

¥ If you’re using LM324s because they’re cheap and
because you have 1/4 LM324just sitting around, consider that
the LM324, because of its class-B output stage, is one of the
worst op amps for a Sallen-Key filter. That’s because its output
impedance is potentially soft.

To minimize this problem, connect a suitable pull-down on the
output to one of the rails, perhaps 15k to -V s (or in some cases to
+V S ). This can cut down on the amount of “feedthrough” by mak-
ingVouT stiffen We have many other amplifiers, and all of them
work better than the LM324.

Which op amps have inherently lower Z 0[jJ , and thus tolerate
current fed in through R1 and Cl? The NSC Webench site at
http://webench.national.com/appinfo/webench/filters/design_r
equirements.cgi does
a surprisingly good job
of predicting how
much the Z 0U t will
hurt, so it’s worth a try.

However, it won’t let
you substitute in differ¬
ent values for the Rs
and Cs, and it won’t let
you break up R1 into
two pieces or add 03.

You’ll have to model that in Spice (which I don’t recommend) or
make a breadboard (which I do recommend), ffi

NEXT PROBLEM ¥ I don’t have any 1600-pF capacitors in my lab.
That’s a mongrel value. So again, we add in another scaling factor
of 1.59. We can now use 1000-pF capacitors and 158k resistors.
Those are easy to find. Thus, you can arbitrarily scale the capaci¬
tors and resistors, up or down, and maintain the same response. I
usually avoid capacitors smaller than 200 pF.

Comments invited! rap@galaxy.nsc.com —or:
Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

ED ONLINE 13480

18 • •

09 . 28.06 ELECTRONIC DESIGN

PeasePORRIDGE

BOB PEASE, Contributing Editor

• • rap@galaxy.nsc.com

What's All This Bicycle Stuff, Anyhow?

friend wrote to me that flying a plane is like riding a
bicycle. It took me about four seconds to rebut him:
“No, flying a plane may be a little similar, but it’s much
different from riding a bike.” A plane can fly fairly

stably with no controls. But a bicycle has to be steered and
controlled so it doesn’t fall over.

And there’s a big difference between flying a plane and rid¬
ing a bike (or motorcycle): If you want a plane to turn, you can
quickly start banking and turning. But to get a bicycle to turn
left, you first have to turn the handlebars to the right and wait
for the bicycle to lean to the left. After that, you can start turn¬
ing left. Motorcycle riders can maneuver faster than cars, but
only if they know what’s coming. Drivers in cars can make
unplanned maneuvers faster than cyclists.

It’s amazing that so many kids learn how to ride a bicy¬
cle. It really is tricky. After learning to ride a bike, driving a
car is easy. That reminds me of some of the early work of
George A. Philbrick. It must have been about 1942, when
the war preparations were ramping up.

George got involved in training anti-aircraft gunners. He
showed that it was quite hard to learn to aim a gun with an
extra lag inserted in the aiming mechanism. But after the
gunner learned how to handle that lag, he became a much
better gunner once the lag was removed. So that was put
into gunners’ training.

Well, riding a bike doesn’t exactly have a lag. But it does
involve a challenging problem—having to steer left before you
can turn right! That’s counter-intuitive. After a kid has learned
to deal with that, driving a car is easy. As I’ve said, I wouldn’t
want to teach a kid to drive a car if he hadn’t already learned
to be a good bicyclist.

I like many things about my new mountain bike, a Special¬
ized Rockhopper A1FS. But I can’t ride it no-hands. It really
has no stability for that. I’m not sure if its good maneuverabili¬
ty is because of this instability or in spite of it. But since I
don’t need to ride no-hands, I don’t complain much. But I
have ridden many miles no-hands on other bikes.

Of course, one operating mode is worse than not using
your hands: If you cross your arms on the handlebars, you will
quickly crash. So don’t do it! If you just touch a finger to the
handlebars and think hard, you might be able to avoid crash¬
ing, but be careful. Assume you will crash anyway.

¥ The recent debate over whether or
not the better tires should be put on the front or the rear of
the car has been settled in favor of the rear. That’s

Jeff Fisher took this photo in June 2002 at 14,600 feet, looking down on
Thorong Phedi (three tiny huts right above RAP’s helmet), just before we got up
to the pass at 17,771 feet. Kalu Tamang, the guy in the picture, was our guide.

because skidding with the rear wheels is so “dangerous.”
Yet we all remember riding our bikes down a slope, locking
up the coaster brake, sliding the rear wheel stably, and
steering just fine!

Remember to steer in the direction of the skid. So, rear-
wheel skids aren’t so bad if we think about it and can steer
quickly. Just don’t overdo it and rip up your tires—or your
father’s lawn.

Our bicycle trek around the Annapurnas in Nepal (never
higher than 17,771 feet) scheduled for last June had to be
postponed. All five of us (even the sherpa) had too much
work and couldn’t get the time off. We’ll try again in May
2007. Check out www.national.com/rap/nepal/index.html
sections 13,14, and 15 if you’re interested in joining us. 1

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090 , Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

18 ••

10 . 12.06 ELECTRONIC DESIGN

PORRinGF

WHAT'S ALL THIS BEST
STUFF, ANYHOW?

Bob Pease

CONTRIBUTING EDITOR

I ’ve been working with
precision op amps.
Some have good linear¬
ity. Some are excellent. Some
have high Zqut* Some have
low. Some are bipolar, oth¬
ers are CMOS. The figure
shows the basic test I’ve
been using for linearity to
exercise the output to 20 V
p-p by applying a ±10-V sine
or triangle wave to the sig¬
nal input.

Meanwhile, the noise gain
of 1000 (R2/R1) magnifies
the input voltage V(e) by a
factor of 1000 so the scope
can see (on a 5-mV scale) a
5-pV signal in cross-plot
mode. The noise can be as
low as 5 pV p-p, and you
can see the distortion of just
a microvolt or two riding
along under the noise.

It’s true that many engi¬
neers are interested in the
total harmonic distortion
plus noise (THD+N). How¬
ever, the distortion riding
along under the noise is
sometimes important, even
though an Audio Precision

can most easily measure
THD+N.

I did some evaluation on
the LM4562 precision audio
op amp. Its noise for an
audio bandwidth is down
near 0.4 pV RM s, and at least
that’s easy to measure. But I
still had to measure the dis¬
tortion at 1 kHz. The test cir¬
cuit in the figure can’t show
the linearity of the gain at 1
kHz, only at 5 or 10 Hz.

That’s because the ac error
is so big, it isn’t easy to see
the distortion. For example,
the LM4562’s ac gain at 1
kHz is about 60,000. Not
bad, but for a full output,
the summing point error is
333 pV, and it’s hard to see
if that is linear within a few
microvolts.

CHEATERS EVENTUALLY
PROSPER • So I decided to
cheat. I used a small vari¬
able capacitor —a few inch¬
es of twisted pair, often
called a “gimmick,” using
teflon wires. I connected
this from the V IN to the

input of the op amp. As I
wound up the wires, the ac
component of the error
voltage shrank a lot.

I kept increasing the fre¬
quency as well as the capaci¬
tance. Finally at 1 kHz, I got
it down under 1 pV of signal
plus a few pV RM s of noise.
This provided good insight
into what the 1-kHz gain
looked like. But why was the
noise so big?

I realized that I’d been
using this lazy man’s gain test
for so long, I wasn’t paying
attention to the way the noise
of the lk resistor (about 4
nV/VHz) was bigger than the
op amp’s noise. So, it was
time to cut the impedance
levels! I didn’t rewire the cir¬
cuit. Instead, I just slapped in
20k across each 1 MO and
20 O across the lk. Of
course, the capacitance had
to be scaled up too, so I put
in about 140 pF on top of the
3-pF gimmick.

This provided a definitely
improved view of the distor¬
tion, with an improved noise
floor. I could see that the ac
distortion, even at 1 kHz,
was somewhere well below
1/2 pV p-p. But I still could¬
n’t see exactly how low. So I

got mad and fed this signal
into our HP3561A spectrum
analyzer. This plainly
showed the amount of the
distortion, such as 71.45 nV
at 2.2 kHz, with a 10k load.
(It degraded to 200 nV with a
lk load.)

The combination of the
subtracting and self-ampli¬
fying effects of my circuit,
plus the ac cancellation, plus
the high resolution of the
spectrum analyzer, showed
-159 dB of distortion at 2.2
kHz (second harmonic)
when running the LM4562
at a 20-V p-p sine output at
1.1 kHz. This was the best
distortion I have ever seen,
and fortunately the best test
circuit I have ever seen, or
we wouldn’t have been able
to measure it.

Comments invited!

rap@galaxy. nsc. com — or:
Mail Stop D2597A, National
Semiconductor
P.O. Box 58090, Santa
Clara, CA 95052-8090

BOB PEASE obtained a BSEE
from MIT in 1961 and is staff sci¬
entist at National Semiconductor
Corp., Santa Clara, Calif.

ED ONLINE 14109

20 ••

12.01.06 ELECTRONIC DESIGN

Pease PORRIDGE

BOB PEASE, Contributing Editor

• • • rap@galaxy.nsc.com

What's All This Floobydust,

Anyhow? (Part 14)

y old friend Robert M. Milne retired a few months
ago as editor of Electronic Design after 23 years.
RMM has been my main contact with ED for 15
years, ever since this column began.

Bob has been very helpful, and a really good sport, about edit¬
ing my columns. He can slice a few words out of a sentence with¬
out wrecking it, or chop a couple of small sentences out of a para¬
graph without changing the meaning—quite an art. I will miss him
a lot. He even tripped me up on technical errors and questioned
me properly. I love the man.

Bob had fairly lousy health in the last few years, but a recent
change of doctor has brought him some good advice: “Cut out
the calcium channel blockers.” Now he is beginning to feel pep¬
py, rather than like a zombie, so he can start enjoying his retire¬
ment. Maybe ducking the need to edit my column on deadline
will even stabilize his blood pressure! Best wishes to you, Bob
Milne. Keep questioning your doctors! Thanks for all the help
and advice over those years.

DOCTORING STOFF, PART 4C STROKE DIAGNOSIS ¥ Many people
know that in case of a heart attack or stroke, it’s very impor¬
tant to get the victim to medical care very quickly, within much
less than an hour. But what do we know about diagnosing such
an unhappy person?

Defining a heart attack relies on asking the victim about various
kinds of pain. Pain in areas near the heart can tell you that you
should call 911. No comments on this today. But in case of a possi¬
ble stroke, where a person may experience numbness or paralysis,
particularly on one side of the body, what can we do as laymen?

First, ask the person to lift both arms and hold them out in
front. Often, a stroke victim will be unable to bring up both
arms or hold them up. They will droop. Serious clue.

Second, ask the person to smile! If the victim can’t smile in a
symmetrical way, you have another serious clue.

Third, ask the person to speak a simple sentence.

I know some people who are barely able to pass some of
these tests in normal situations! But barring that, call 911 and
describe these serious symptoms. The dispatcher usually will
send an ambulance promptly.

Treating a stroke within the first hour gives the victim a good
chance to recover. Somebody who trips or starts mumbling
(worse than usual) may be okay. But if you aren’t sure, ask these
questions, call 911, and explain how the person flunked these

tests. The 911 dispatcher may ask other questions, like those at
www.strokecenter. org/education/ja uch/02. h tm#Chain.

After all, every one of us is getting older by definition. And as
all of our friends get older, the likelihood of somebody having a
stroke increases.

RELATED TOPIC ¥ If you have medicines past their expiration
date, please do not flush them down the toilet. Flushed medi¬
cines have been found to contaminate our waterways, leading
to fish with hormonal afflictions, schizophrenia, and unhappy
sex lives. Save those medications in a safe, locked place until a
pharmacist or doctor tells you how to dispose of them safely. I
keep most of my old medications in my freezer, so the effective
expiration date is extended (except for aspirin, whose expira¬
tion apparently isn’t extended by freezing). ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

ED ONLINE 14195

16 -

12.15.06 ELECTRONIC DESIGN

Pease Porridge

BOB PEASE | CONTRIBUTING EDITOR

What’s All This
GetHuman Stuff,
Anyhow?

e ordered a new refrigerator from Sears to replace our sick, dying 30-year-
old icebox, and it was supposed to be delivered on December 7. The Sears
salesman told us that we a get a phone call about delivery on December 6.
When Sears called, the operator said it would be delivered the next day

between 10 a.m. and noon. But the next morning,
another operator called to say the refrigerator
hadn’t (magically?) arrived from the man
ufacturer, so we should wait a day or
two. So much for JIT...

Since our icebox didn’t come on
Thursday, my wife Nancy waited
most of Friday to see if Sears would
call her with news about delivery.

Finally, deciding to be proactive, she
called Sears.

The first person she talked to tried
to give her a runaround: “Oh, that
was delivered on Thursday...” Nancy
explained that no, it obviously was
not delivered on Thursday (stupid
computers...). The Sears person
said someone “would call back in
24 to 48 hours” to tell her when it
could be delivered.

Nancy signed off, and then she
gradually began to fume. She finally
followed the advice of a friend,

regarding an airline that gave her some stupid advice: “Call
the airline back again, and maybe you can talk to a different
person who can give you a more reasonable answer—or an
answer that you like.”

(She once had called an airline whose clerk had told her it
was impossible to change a ticket because it was a paper tick¬
et, not an E-ticket.)

This is contrary to the recent saying that insanity consists of
asking the same question over and over and expecting a differ¬
ent answer each time you ask it.

Recently, I’ve been asking my computer to do a simple
task, and it refused to do it. So I would reboot, but I got no
improvement. I would reboot one more time to finally get a
reasonable response. Sometimes... maybe... sometimes, two
or three more reboots.

So Nancy got back on line with a person at Sears who con¬
sulted her computer and said, “Oh, yes, the refrigerator will be
delivered on December 10.” Nancy queried, “But that is a Sun¬
day. Do you deliver on Sundays?” Oh, yes.

So it goes to show that when you’re
talking to a real person who is address¬
ing a computer, you may be talking to
the wrong person and getting the wrong
answer. Sometimes, trying again is the
right thing to do. (Meanwhile, we gotta
give those delivery guys a nice tip for get¬
ting it delivered as fast as possible.)

Get Out Of The Phone Tree • There’s
a list of at least 475 major U.S. companies at
the “gethuman 500 database” at www.geth-
uman.com. You can get simple instructions
on how to talk to a real person at these real
companies without wasting dozens of min¬
utes on a dumb “phone tree.”

Sometimes the instructions are as sim¬
ple as “press 0” or “press # at each
prompt” or even “don’t press anything,
don’t say anything.” It varies from com¬
pany to company.

I forecast that the “future” doesn’t
belong to computerized phone trees,
except for the simplest information, or to “outsourced” help
centers who try to “help us” from 6000 miles away. The
“future” belongs to wise humans who can answer a phone,
help us solve problems, and save us time.

It belongs to humans who can ask us the right questions so
we can resolve the real problem quickly with a good answer,
like “the refrigerator will be delivered between 3:30 and 5:30
p.m. on Sunday.” Would you believe that? (And yes, it was.)

When people call or e-mail me, I do try to be very helpful
or at least transfer them to someone who can be helpful. I
rest my case. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

Bob Pease obtained a BSEE from MIT in 1961 and is staff scientist at
National Semiconductor Corp., Santa Clara, Calif.

ED ONLINE 14414

01.11.07 ELECTRONIC DESIGN

ED ONLINE 14550

BOB PEASE | CONTRIBUTING EDITOR

rap@galaxy.nsc.com

Bob's Mailbox

i Bob: I was just reading your column in the Nov. 16,

2006 issue of Electronic Design (“What’s All This Si0 2
Stuff ; Anyhow?” p. 18) and had some insights. The silicon
dioxide (Si0 2 ) added to foods for anti-caking is a little bit
different from sand. Chemically, it’s the same, but it’s made by dis¬
solving Si0 2 in some really nasty acids and then flashing the acids
off. What’s left looks like a semi-solid fog. My old boss used to joke
that it has “negative weight.” You could keep scooping the stuff
onto a scale for quite some time before you got much of any reading
at all. There was no resistance to the scoop. It was like scooping fog
out of a bottle. I think the trade name was CabOSil. When you blend
it in with other powders, it keeps them from caking without adding
really any weight. It also acts as a great thickening agent in oils and
sauces or soups. Who would have guessed? (I often put a little roux
of butter, flour ; and cream into my recipes as a thickener ; but the
Si0 2 foam would have fewer calories. On the other hand, the roux
does add to the flavor./rap) As you pointed out in your column, Si0 2
is a marvelous material and is very useful in its many forms.

• Steve Krueger

• Pease: It sure does help maintain our beaches! Where
would our seashores be without it?

Hi Bob: Regarding your article on Si0 2 —wow, finally. It has inter¬
ested me that in the areas of food, drugs, and electronics, the role
of passive components often is ignored. If something is nutritive
or electrically conductive, volumes are written. Yet the passive
components often are either barely mentioned or ignored. (We
can stop worrying about Si0 2 because it is soooo good!/rap)
Almost everything electronic—ICs, transistors, capacitors, resis¬
tors, battery cells, whatever—has a certain amount of conductors.
Yet none of these would work without non-conductors (insulators)
or at least dielectrics. (For sure, and Si0 2 is one of the best. It has
very low leakage when used to isolate input nodes, often better
than 10 15 Q, perhaps better than 10 20 Cl-cm—low leakage when
used in capacitors and low dielectric absorption too! Maybe not
quite as good as air.../rap) In most (all?) electronic components,
insulating material occupies more space, or area, or volume than
conductors. (Maybe not in rectifiers.../rap) One of the most com¬
mon non-conductors (or low conductors) is Si0 2 . Although sand is
rarely pure Si0 2 in the natural state (most sand grains also con¬
tain a wide variety of other elements and compounds, such as the
conductor aluminum), Si0 2 is probably the most common major
component. Yet slightly inland Silicon Valley isn’t called that
because of its beaches (or relative lack thereof). Your column
mentions silicon as one of the components mentioned in the
ingredients of various food products. This, of course, is an FDA
requirement. Golly gee, there doesn’t appear to be any govern¬
ment organization that requires such detailed listing of the com¬
ponents used in transistors, etc. (Such as arsenic.../rap) As long

as you don’t eat it, any such listing is entirely voluntary on the part
of the manufacturer. When people start eating ICs, this situation
may change. Chickens frequently eat sand. Si0 2 has little, if any,
nutritive value. But chickens don’t have any teeth. Sand grains
serve as an abrasive material that breaks down other things they
eat. Although Si0 2 passes through their bodies almost
unchanged, it serves a useful mechanical function. (When I was a
kid, we fed our chickens calcium carbonate in the form of
crushed shell fragments. That was designed to work much better
than sand./rap) This whole subject really should be explored
more thoroughly, especially in regard to electronics. Certainly,
insulators (those things that do conduct electricity, but just bare¬
ly), like air, are very useful. If you’re working with very sensitive
components or transmission items, like video, you don’t want to
wrap your ac lines around video cables, even if they’re shielded,
securely grounded, and insulated with a half-inch of rubber. (Next
time I am running video, I will wrap some ac power lines around
them and see if I can tell any difference, /rap) Yeah, a sandbox
can do surprising wonders, but even this is imperfect.

• John Rupkalvis

• Pease: Thanks for the comments. Best regards.

Hello Bob: Your columns are sometimes amusing, but when will
you finally switch to the metric system as used throughout
Europe? (Maybe when we get worldwide cooling. Do you happen
to know the schedule for Hell freezing over? That is about when I
will change over entirely. When driving in Germany or France, I
think in kilometers and kilometers per hour. But even there when
somebody asks me “How long is your foot?” I say it is about 1
foot long, or 1.1 feet with my shoe on. The reason we think in
feet is because a foot is a useful measure./rap) Your columns
about your bicycle treks only use feet. Do you prefer to use old
units in electronics as well? (I could use AB-volts and STAT-volts,
but those are a little too obscure. But I suppose I could start
using them.../rap) How old-minded would that be?

• Istvan Cocron

• Pease: About 65 years. Even in England, many older people
don't like using the metric system. An old man in England
wouldn't buy a kilogram of meat or half a kilogram. He would
buy 454 grams because that is what he needs. And he would
certainly buy a pint of beer. When Hell freezes over, and the
beer does too, then we'll consider changing. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

01.18.07 ELECTRONIC DESIGN

ED ONLINE 14701

BOB PEASE | CONTRIBUTING EDITOR

rap@galaxy.nsc.com

What's All This Diffusion Stuff,
Anyhow?

nee upon a time, if you wanted to make a junction transistor, you could
start with a small seed crystal of pure germanium. Using Czochralski's
1917 methods, the crystal was gradually rotated and pulled out of a lightly
doped N-type material and grown into a small boule. After suitable growth,

a small amount of acceptor impurity such as gallium was added to
the melt, causing the germanium to form a PN junction. This was
all consistent with the theory of William Shockley, published in
June 1949.

The making of simple PN junctions was not that new. Then,
Morgan Sparks of Bell Labs added a stronger impurity (antimony)
to make a second closely spaced junction
and keep the crystal growing. On April 12,

1950, the first junction (NPN) transistor
was born, and all hell broke loose. Hun¬
dreds of inventions were added to make
better transistors.

But this grown-junction transistor was
hard to make, as the base area was barely
25 pm thick and hard to connect to. Many
efforts, both theoretical and practical,
went into making practical transistors. Yet
fora longtime, the performance was still
lousy, and the yields were still poor—a p of
40, f(a) of 15 MHz, and breakdown of 25
V was considered amazing.

SOME WINTER READING ¥ Bo

Lojek, a research scientist at Atmel’s Colorado Springs facility,
chronicles the amazing stories of all kinds of brilliant research in
his new book, History of Semiconductor Engineering. He also
documents all of the human foibles that mark the industry’s
milestones.

Shockley, the “Fairchild Eight,” and Texas Instruments are just
some of the major players Bo describes. Jean Hoerni’s planar
process was a major advance, forcing the junctions to be made
under a silicon-dioxide layer and greatly improving reliability under
severe temperature conditions.

Dozens of minor companies made little advances. Some of
these advances were lost forever, mere speedbumps on the way
to progress. Dozens of engineers and scientists contributed ideas
of varying degrees of helpfulness or uselessness.

The diffusion of gaseous impurities into a silicon wafer, masked
by wax (or later by silicon dioxide), was a major factor in these
improvements. Bo even includes the diffusion formulae on how
you could make your own transistors, just as the original Fairchild
researchers did. Planar process? Do it yourself!

Unfortunately, this book neglects to mention the pair of small,
tasteful monuments to the contributions of Bob Widlar and Jean
Hoerni at the foot of San Gabriel Court in Sunnyvale, right off Kifer
Road, near Maxim’s headquarters. I think Bo was too modest to
mention that he had instigated the building of these monuments
to the pioneers of our industry.

The bold Robert J. Widlar gets his own
chapter and a half, showing how the
“champion’s” approach to pioneering
improved (linear) circuits at Fairchild and
later at National. His collaboration with
Dave Talbert, who optimized the diffusion
processes (in conjunction with Widlar’s
needs), is well documented.

When you’re a pioneer, you may have to
work extra hard to avoid arrows in your
back and prove your ideas will really work.
Widlar did that.

The book includes many drawings of
classic inventions and photos of great peo¬
ple (and sheep), illustrating the stories. It
also describes the business practices and
human foibles that led to success (or fail¬
ure), as well as the full panoply of human brilliance and stupidity.

PICK UP A COPY ¥ Where can you get your own History of
Semiconductor Engineering ? Go to www.amazon.com and just
punch in Bo Lojek. There are some good comments and reviews
in there. Or, go to www.electronicdesign.com and enter ED
Online 14727.

And in the interests of full disclosure, I helped proofread his
book and offered many corrections because I was entranced by
his storytelling. Bo has a full 6-ft file on Shockley and many dozens
of feet of files on the other players. He distilled the best stories
down into this 380-page book. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE brom MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

ill' »

fjj * f

■ “ * ~

fiiflt EUli

Circuit by Bob Widlar; layout by Dolores Talbert;
book by Bo Lojek

02.01.07 ELECTRONIC DESIGN

ED ONLINE 14762

BOB PEASE | CONTRIBUTING EDITOR

rap@galaxy.nsc.com

's Mailbox

ello Mr. Pease: I’ve read somewhere that it can be a
problem for the op amp to lose one of the supplies or
use the wrong startup sequence. (You are correct A
startup sequence can cause great trouble on the posi¬
tive and negative rails, /rap) Is it always a particular rail? Can you
shed some light on the cause of this? Does it apply for bipolar
and CMOS families?

• David Smith

• Pease: You have proposed a tough problem, and I don't
think there are any simple answers because there are only
a couple thousand kinds of op amps that people use and a
few dozen configurations for each. What a mess! Let's see if
I can provide a general solution.

(A) Please try to avoid nailing the positive input of an op amp
to ground, even if it should be "grounded." Bob Widlar has
argued that even lk is a good idea to prevent gross input
currents from flowing.

(B) Here is a worst case. Set up a nice, healthy analog sys¬
tem with nice, healthy power supplies—say, ±15 Vat 1 A.
Connect an op amp as load with its positive input grounded
and with 30 0 of load from the +15-V bus to the -15-V bus.
Now to get in trouble, connect the ground and the positive
rail. Refuse to connect the -15-V supply to the -15-V load.
The 30-0 load would try to pull the negative power-supply
bus toward +13 V and feed 480 mA through the 30-0 resis¬
tor to pull the op amp's negative rail above its positive
input—not a good idea. Many op amps will blow up quickly if
you try this. LF356 biFETs and LM324s will probably blow or
be badly damaged. Many bipolar and CMOS op amps, too.

(C) A partial solution is to take the load (the operating circuit
and system) and connect anti-reversal diodes across the
power-supply busses of the load (the system) so the nega¬
tive supply can go barely 0.8 V above the ground bus and
the positive supply can go barely 0.8 V below the ground.
1N4004S would be good in most cases. I don't think you will
need 1N5819S.

(D) Avoid connecting a wire short from the input of the op
amp to ground. If you put at least lk there, it will not add
much noise, but it will limit the currents and prevent
much harm.

Robert: Thanks so much for speaking at the Audio Engineering
Society conference in San Francisco in October. I was wondering
how PPS film capacitors such as Panasonic’s ECH-U series com¬
pare with the rest in your testing. (Polypropylenes have had supe¬
rior soakage (dielectric absorption) and superior leakage (Tau =
several years)/rap) I’m in an application where I have to use SMT,
and I’ve heard that PPS does even better than NPO for audio sig¬
nal paths. (If you can use a big enough poly it can sound better. If

you use NPO and you chintz because it's not big enough in C val¬
ue, you can't blame it on the inherent quality of the dielectric.
Many people use poly and other films where SMT is "required"
because they are superior and worth it in terms of performance.
Solder them in by hand. Have fun!/rap)

• Tyler Gleghorn

• Pease: Why are the PPS capacitors so good? Because they
are polypropylene, which is a very superior dielectric. Some
people think it's an octave worse than polystyrene. I think it
is an octave better, for all the applications I have seen.

There are many manufacturers of polypropylene and very
few now of polystyrene. Polypropylene is comparable to NPO
for audio work, but you can get 1 p For more in a reasonable
package, whereas NPO/COG is only available up to 0.1 pF.
However, many film capacitors such as poly aren't very
available in surface-mount (SMT) format. NPO/COG is much
better than poly for ternpco, though for audio circuits, who's
going to complain about -120 ppm/°C versus (0 ±30)? If
you think you can hear any difference, I suggest you try
some A-B-X testing. I don't think anybody can hear any dif¬
ference in an audio circuit.

Bob: I finally got to read your Nov. 6, 2006 article ("Bob's Mail¬
box," p. 18). I thought John Cook’s comment on the two-chassis
RCA studio amplifier was interesting. I’m sure that separating it
into two chassis, one with the power supply, reduced hum due to
magnetic coupling from the power transformers. (Not to mention
the transients when the rectifiers carry big surges of current for
short times./rap) Years ago, my company made audio mixers for
the television broadcast industry. We had mu metal shields on
our microphone input transformers and a steel shield around our
power transformer. The noise was great, until you brought up sev¬
eral channels. (Ha!/rap) The fix, developed by one of my part¬
ners, was to reverse both primary and secondary windings on
every other channel. Phase was maintained through the trans¬
formers, since both were reversed. But as more channels were
brought up, the magnetically coupled hum went down instead of
up. (And this was the input transformers (not the power trans¬
formers)? What a brilliant solution! I don't expect things like that
to cancel out./rap) I always thought it was a clever cure!

• Harold Hallikainen

• Pease: Yeah, fiendishly... Thanks for the comments, ffi

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

02.15.07 ELECTRONIC DESIGN

BOB PEASE | CONTRIBUTING EDITOR

rap@galaxy.nsc.com

What's All This
Lighted" Stuff,

"Others Stay
Anyhow?

he box of Christmas lights said "If one or more lights go out, others stay light¬
ed." Yeah, sure. My wife bought several cheap boxes of 1 00-light strings,
and they looked very nice. This year, she had great plans to drape them
along a fence, which would look real pretty. But some of the strings were

dead—kind of frustrating when you want to put the lights up now.

I did some simple checks. If one bulb went open, the others in
series went out. Yeah, 5-V double-anode zeners can keep the oth¬
er lamps running, but that’s not what you get when you pay $1.06
for 100 bulbs. As I hadn’t planned to do any troubleshooting, I
had no tools or meters with me—not even an LED.

To avoid disappointing my wife, I had to get to work and fix
them, that night, so she could impress some friends who were
coming over. It wasn’t a question of having spare bulbs. I had
plenty. But I really didn’t want to pull out all of the bulbs and test
them, as that surely would have done more harm than good.

The bulbs’ bases were flaky too, and unplugging and plug¬
ging would have ruined them. I didn’t even trust the sockets! I
finally found some lamp cord from an old lamp, a couple of
nails, and some tape for insulation.

TO THE TEST • These 2.5-V bulbs would take a little overvolt¬
age testing. So I pulled out the first and eleventh bulbs and
shorted out that group with the wire and nails. I decided there
was no lack of continuity. I went on to the group of 11 to 21, 21
to 31, and 31 to 41—still no effect. Oh, what I would have given
for an ohmmeter or a voltmeter!

Finally, I shorted out bulbs 41 to 50, and the other 40 bulbs
lit up nice and bright! Hooray. As I indicated on page 7 of my
book on troubleshooting, “troubleshooting a series divider is
sometimes a nice challenge”—yeah, if you can gain access to
the resistor string, which wasn’t easy in this case. I suppose I
could have taken a pin and poked it through the insulation into
the wires, but that’s not so great.

I checked continuity between 41 and 45 and between 45 and
50, and both were open. I had at least two failures, so I had to be
pretty careful making assumptions about binary searching! By
this time, some of the bulbs were getting flaky. I set aside some
bulbs as “not trusted” and threw out a couple as hopeless.

One bulb refused to light but apparently would pass current.
Another bulb seemed burned. And, one bulb had a melted
base. But finally I got all the strings working with minimum
tools—just a wire and some nails.

I’m going to use a tiny test kit in my future Christmas decorating
chores, probably a 4.7k 2-W resistor and/or a 1-pF 200-V capaci¬
tor, as well as a couple of insulated alligator clips and safety pins.
Meanwhile, trusting cheap foreign junk isn’t always a good idea.

I’m not really mad at the bulbs’ manufacturer. I just wish it was a
little more truthful about the probable failure modes.

CLASSIC LIGHT BULB TRICK • Let’s say you have a motor-
generator set (hydro power?) to generate power and send it
back into the 60-Hzgrid. But you can’t just connect your gener¬
ator, as it’s likely to be out of phase. Connecting it could blow a
fuse or cause damage. What to do?

Tie the common of your generator to the grid and connect a
couple of 120-V incandescent bulbs in series from your output
to the grid. Adjust your power input to get your frequency close
to 60 Hz, and then bring the phase as close as possible.

When you see the intensity of those two lights drop gradually
to zero, you know the frequency is very close and the phase is
very small, so you can throw the switch! Of course, this trick is
only a little over 110 years old. No fancy phase meter is need¬
ed. This trick does the job. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

03.01.07 ELECTRONIC DESIGN

. Peas

Porridge

BOB PEASE I CONTRIBUTING EDITOR

ED ONLINE 14972

rap@galaxy.nsc.com

Bob's Mailbox

i Bob: I have been following your silicon-dioxide (Si0 2 )
articles and just saw Steve Krueger’s response (“Bob's
Mailbox/' Jan. 18, p. 16) naming it under the trade
name Cabosil. We use Cabosil at our company as a
thickening agent for adhesives and epoxies during manufacture.
When I had my first exposure to it, I was told it was small glass
spheres. (There are some epoxies with air-filled glass spheres,
but that's different. Those are usually to yield lower dielectric
constant and less capacitance, /rap) I was concerned about the
size of the “spheres” (breathing them in) and did some research.

I found out it is the thickening agent for ketchup—the reason we
had to go to squeeze bottles versus glass bottles (not an excess
of tomatoes)? (I have the impression that there is a lot of vari¬
ance in the natural thickness of tomatoes or ketchup. My wife
solves this by simmering until the texture is right, but Del Monte
ain't gonna waste time like that!/rap) I was shocked about glass
or sand added but still love my ketchup.

• Kevin Wagner

• Pease: Check! Best regards.

Hi Bob: We are hoping you can shed some light on the issue of
capacitor soakage (or dielectric absorption, or “DA"/rap). We
have read your publication on the subject entitled “Understanding
Capacitor Soakage to Optimize Analog Systems” (see www.nation-
al.com/rap/Application/0,.1570,28,OO.html). It was very helpful
in understanding DA. We are currently struggling with an integra¬
tor that requires a reset to zero. After the reset, the integrator
slowly goes in the direction of the previous output. This integrator
is used in a closed-loop feedback circuit to create a very low-fre¬
quency high pass filter (0.01 Hz). We are using 1 pF and 3.6 MQ
as the current RC values. After studying your article and trying
many different dielectrics (polypropylene, PPS, polyester) (First of
all, get the “polyester" or mylars out of there. They are at least 10
times inferior./rap) we have found with all the capacitors we have
tested that the internal soakage elements have a substantially
longer time constant than what you have shown in your article.

• David Muir

• Pease: All capacitors—the ones I measured and the ones
you measured—have many more time constants. You are
quite right. I never intended Figure 4 in my article to cover
all of the time constants—just the ones relative to 5-second
shorting periods. I shoulda indicated with dashed lines that
there are more Rs and Cs out there. On and on...

I do not know which time constants will change (nor how
much) as you short out the cap for longer times. Bigger Cs?
How much? I can't guess. You will have to measure the actu¬
al data on your favorite caps.

I know that on my favorite polypropylenes, as the time peri¬
od goes out into days and weeks, if you let the soakage set¬

tle, it seems to subside into months. The observed leakage
rates (half of which are probably soakage) goes out into
years! No kidding!

Hi Bob: I read with interest your “Mailbox” in the Jan. 18 issue,
specifically the exchange concerning the metric system. You say
“Even in England, many older people don’t like using the metric
system.” I would say this is true even of younger generations
here in the U.K. Although we are supposed to be metric, in keep¬
ing with the rest of Europe, the reality is that imperial measures
persist in many areas of life. Typical examples of such measures
would be miles for distances, miles per hour for speed, gallons
for petrol, pints for beer, and pounds, ounces, and stones for
weight. There have even been instances where market traders
have faced prosecution for selling goods in imperial measures (to
much public outcry, I might add). Imagine winding up in court for
selling two pounds of bananas! (instead of 0.9 kg... /rap) The
mind boggles. It seems to me, though, that the approach manu¬
facturers have taken, particularly in relation to foodstuffs, is to
sell things in imperial quantities but label them with the appropri¬
ate metric measure. So for instance, you can still buy a pint of
milk but it will be labeled as 0.45 litres (or whatever the correct
conversion is). (Otherwise, you'd have to throw out a lot of pint
bottles.../rap) For my own part, I tend to use metric measures in
my engineering life (simply because it’s easier) and mainly impe¬
rial measures for everything else. I went to school at a time when
the U.K. was just starting to move from one system to the other,
and so I’m somewhat caught between the two stools, so to
speak. If you asked me to imagine walking a mile or lifting 10
pounds, I could do so quite easily, but the same would not be
true of walking a kilometre or lifting a kilogram. Also, if you were
to ask someone over here how far it is to the next town, I think it
is almost inconceivable that they would give you an answer in
kilometres rather than miles. (If the guy who asked me had
French license plates, I could do it.../rap) So, imperial measures
are definitely not dead over here. We Brits value our independ¬
ence. (Bravo! And we too, until Hell freezes over ; which could
happen any week now. /rap) And if you want further evidence of
that fact, you should see how strenuously most of us are oppos¬
ing the Euro!

• Christopher Hill

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

03.15.07 ELECTRONIC DESIGN

ED ONLINE 15116

BOB PEASE I CONTRIBUTING EDITOR

rap@galaxy.nsc.com

What's All This Capacitor
Leakage Stuff, Anyhow?

e all know that capacitors have a shunt resistance (leakage) and that
leakage resistance should be pretty easy to measure, right? Wrong! I've
measured a lot of capacitors for short-term soakage (dielectric absorp¬
tion) per www.national.com/rap/Application/O, 1570,28,00.html.

After the short-term soakage stops, it’s possible (not easy) to
measure the leakage. For example, if you charge a good cap up
to 9 V for a few seconds, it will start discharging shortly for sev¬
eral millivolts. If you wait long enough, you may see leakage
slow down to a few millivolts per hour. But you will see the long¬
term soakage. Is that different from the short-time leakage?
Maybe not.

Now I will charge up some of my favorite low-leakage
capacitors (such as Panasonic polypropylene 1 pF) up to
9.021 V dc (a random voltage) for an hour. I will read the
V 0 ut with my favorite high-input-impedance unity-gain follow¬
er (LMC662, lb about 0.003 pA) and buffer that into my
favorite six-digit digital voltmeter (DVM) (Agilent/FIP34401A)
and monitor the V 0 ut once a day for several days.

Why did I choose 9 V? Because that’s within the common¬
mode range of the op amp and the DVM at highest resolution. I
keep the input ball hook connected to +8.8 V dc between read¬
ings. I also keep my left hand grounded to +8.8 V.

DAY BY DAY ¥ One of my e-mail colleagues had been monitor¬
ing some good 0.1-pF polystyrenes, and he was impressed that
they got down to a leak rate of better than a year after several
months. Well, I could see that my polypropylenes had their leak
rate improve even better than that in just a few days. Refer to the
list of voltages below:

DayO: 9.0214 V
Day 1: 9.01870 V
Day 2: 9.01756 V
Day 6: 9.0135 V
Day 7: 9.0123 V
Day 8: 9.01018 V
Day 9: 9.00941V
Day 11: 9.00788 V
Day 12: 9.00544 V
Day 13: 9.00422 V

The first day after soaking for an hour, their leak rate was as
good as 2.7 mV per day. Not bad.

If you had a 1 million-MQ resistor across a 1-pF capacitor at the
9-V level, it would draw 9 pA, which would pull down the capacitor
778 mV per day. All the capacitor types I tested were better than

this, except some “oil-and-paper” caps that supposedly had spe¬
cial qualities for audio signals.

If you had a 10-meg-meg resistance, that would cause the cap
to leak down 78 mV/day. With 100 meg-megs, it would be 7.8
mV per day. Several good capacitors soon began to leak slower
than that. After a mere week, some of the best caps were leak¬
ing at a rate down near 1 mV/day. Quite good. So, what’s the
big deal?

The big deal is that a time constant of 31.5 meg seconds is one
year! So any capacitor leaking less than 2.5 mV per day is leaking
at a tau (rate) of 10 years or more. If you had to wait a few months
to get this leak rate, well, that’s not bad. But achieving this leak
rate in less than two weeks is, I would say, quite good. Less than a
day? Spectacular.

So I’m finding that good polypropylene caps are better than
the best (old) polystyrenes, in terms of soakage or dielectric
absorption (early or late) and in terms of leakage, early or late.
Are Teflons any better? Not much. I may have to buy a couple to
find out. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California

03.29.07 ELECTRONIC DESIGN

ED ONLINE 15199

BOB PEASE | Contributing Editor

rap@galaxy.nsc.com

Bob's Mailbox

i Bob: Your tip on howto phase a generator into the elec¬
tric grid would work (“What’s All This 'Others Stay Lighted’
Stuff ; Anyhow?” March 1, p. 20). Phasing sets of the not
too distant past used a couple of incandescent bulbs and
a guy poised with his finger on the button, just as you said. I wanted
to explain, however, that in the electric power industry, we go to
great pains to discourage people from connecting a generator into
their house wiring because of the potential safety threat it repre¬
sents. Our linemen goto work on circuits expecting (but not trust¬
ing—more on that in a minute) that the area with the outage has
only one source of feed—from the electric grid. Once they detect any
other source, all repair work comes to an immediate halt while they
troubleshoot it back to that source. (Okay we sure agree. A bus that
connects to the power grid must not be corrupted by a second,
uncontrolled power source. But in the rapidly growing number of
cases where people want to sell power back into the grid (and make
the power meter “spin backwards”), you have to plan to get things
disconnected! Obviously, an “inverter” that’s trying to put a kilowatt
back into the grid at 220 V will not be happy if the grid loses power
and turns into 0.01 Q to ground!/rap) That is what I meant by not
trusting. Our work rules require testing it dead, sectionalizing, and
then grounding prior to starting work. If they follow all those rules
religiously, they should stay safe. (We agree. Safety first. Gotta have
a high-reliability shorting bar!/rap) If they fail to follow all the rules,
and they get hurt or killed by the unexpected source, then the
homeowner has tragically caused the accident. For engineered sys¬
tems where the homeowner is contracting to supply the grid via
photovoltaics, wind, etc., we require installation of relays to detect a
problem on the grid and automatically open an isolating breaker.

(So, these relays and breakers open automatically and then require
good planning before a decision to reconnect?/rap) It would be
unusual for a generator to stay online during a widespread outage.
But it is not impossible, and in almost 30 years, I have seen two or
three cases where it did exactly that. The situation has to be just
right (or just wrong!/rap), but it can happen. I hope your readers do
not attempt to connect their generators to the grid. Maybe you can
advise against it in a future column.

• Gerry A. Akin

• Pease: I guess that’s a fair reminder. We need reliable
relays, and plans, to disconnect any local power sources per
strict rules that didn’t exist 100 years ago. You are right. Pro¬
fessional installations have these relays and plans. Thanks
for reminding us.

Hi Bob: Streetlights were once connected in series too, with an
automatic shunting bar in case a lamp failed open. See the histo¬
ry section at http://en.wikipedia.org/wiki/Street_light. (Thanks
for the historical info! It’s kind of odd to think of an “anti-fuse” that
goes short ifovervoltaged!/rap) See the brochure for constant-

current transformers (to feed such series strings) at www.special-
tytransformers.com/ST%20Acrobat/CCR%20BROCHURE.pdf
(That’s quite weird! A transformer with moving parts! I guess fact
is much stranger than fiction./rap) and government spec (p. 11)
at www.wbdg.org/ccb/D0D/UFGS/UFGS%2026%2056%2019.
00%2040.pdf. For more on Christmas lights, see www.plan-
etchristmas.com/Minis.htm, especially the update at the bottom.

• Jim Harman

• Pease: This does seem to be a complete story about “self¬
repairing” Christmas lights. It seems as if the “SEMI-conduct-
ing” path to the wire has to be pretty reliable or it would short
out and steal all the current away from the actual incandes¬
cent filament! I guess I gotta go connect up several examples
of dead bulbs to a curve tracer to see what kind of non-linear
stuff is in there. I wonder what ohms we will find...

Hi Bob: In the Feb. 15 edition, you refer to PPS capacitors as
polypropylene. While I share your appreciation for polypropylene
as ideal for audio applications, PPS refers to polyphenylene sul¬
fide, a cat of quite a different color. (I had not appreciated the
PPS material. I’ve never worked with it. /rap) PPS capacitors are
available in surface-mount packages and are alleged to survive
typical reflow processes satisfactorily. PPS as a high-temperature
dielectric has a lot of appeal but unfortunately is very expensive
and does not self-heal very well. We have at my company decid¬
ed not to pursue manufacture of capacitors using this dielectric. I
have no idea as to “sonic characteristics” of PPS capacitors. In
spite of the fact that I play lead guitar in a country rock band with
homemade (sand state) gear, I don’t think that there is much, if
any, difference in sound with various film capacitor dielectrics. (I
tend to agree. In any well-designed audio equipment, all films will
work pretty well.) I try to avoid discussions on the sonic differ¬
ences of different capacitor types. Those discussions have often
resulted in “flame-o-grams” from so-called “gurus,” and I no
longer try to add value to related newsgroups. It’s definitely a
“reader beware” environment on the Internet as far as audio and
electronic music equipment is concerned.

• Terry Hosking

• Pease: It is also a difficult place to get any honest compar¬
isons with any two kinds of audio equipment. Bringing up the
concept ofA-B-X comparison is just too horrifying for some of
these grossly opinionated people. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

04.12.07 ELECTRONIC DESIGN

••

BOB PEASE | Contributing Editor
rap@galaxy.nsc.com

What's All This Cold Toes
Stuff, Anyhow?

sure walked into it. I've always known that I can stuff my warm feet (with warm
socks) into frozen boots and just start walking, and they would warm up nicely. These
are Vasque trekking boots, weighing about 2.1 lb each, well insu ated down to
about -20°F and extremely comfortable. So I put them on and hiked up the trail,

all very cozy, on Jan. 17. The air
temperature was around 10°F
to20°F-not bad.

But after several hours, I got
tired and needed some rest. My
metabolism cooled down, my
body and limbs cooled down,
and my feet cooled down—and
they didn’t complain. It’s like
throwing a frog into warm water
and adding heat gradually to
boil the frog. So my feet got
quite cold, and I neglected to
dive into my bivouac sack.

Four toes got frozen badly, and six just frosted. But the next
day, I walked three miles down the hill, very comfortably. After I
came off the hill, I took a shower in my motel room, and my feet
didn’t hurt. I washed them, dried them, and didn’t even look at
them, as they didn’t hurt, and I had no clue they had been
damaged. Four days later, I looked at my toes, and I figured it
out. They looked lousy, and I went to a podiatrist. I’ll save 92%
of my toes.

I’ve learned that I can’t trust my nerves to tell me that my
feet are cold. Dr. Bolognini calls this “neuropathy.” People with
diabetes have to watch out for this. I learned the hard way.

SENSORS TO THE RESCUE ¥ How can I go trekking or hiking
again next year after I get my feet repaired without fear of chilling
my toes? I suddenly began planning (at midnight in my bed) some
sensors to keep an eye on my feet and toes.

I could go visit my Friendly Neighborhood Applications Engineer
for Temp Sensors. And who is that? R.A. Pease, plus Emmy Den¬
ton. So I’m going to write down my plans, show them to Emmy for
a sanity check, and build up some toe temp sensors. I don’t think
it’s going to be that hard.

Should I use the LM35Z, or the LM45M, or the LM62M? I’ll
debate this—any of them would work well. But I’ll bring three thin
wires up my leg (secured using paper tape) and right past my tum¬
my to plug in to a control panel on my chest, hooked on a lanyard
around my neck.

When I push a button, I’ll supply+5 V to the+V S pin of the ana¬
log temp sensor, which will indicate how cold my toes are on a tiny

PE TF^TPR,

3 lock

analog voltmeter. I can monitor my toe temperatures—and fingers,
too, if I want. Will I add an automatic timer to take a reading every
10 minutes? And beep three times if the toes are okay but only
once if they’re too cold? Oh, probably.

Battery life should be several days. I’ll start with four AAA
cells, but later I’ll probably go to a lithium cell. Sensor weight
plus 5 ft of wires should be less than half an ounce, so I’ll be
able to bring spares.

It will be easy to build this sensor, but I can’t very well test it out
for a while. (Oh, I can test it on my wife’s feet!) Will it work if my
boot gets full of cold water? Steam? To be seen. What would I do if
my feet are too cold? Add some foot-warmers. Get in my sleeping
bag. Take heroic measures, as after your toes have been frostbit¬
ten, they are more susceptible to the cold.

The adjacent block diagram shows a simple scheme to apply
power to the LM35 sensors, automatically, alternately, every few
minutes. Then, the two sensors have their outputs paralleled (the
one that’s not powered has no effect on the other sensor) and fed
to a voltage-controlled oscillator, so we can hear who’s cold. The
resistor-diode path tells us the power supply is alive. For the com¬
plete schematic, see www.national.com/rap/coldtoes.html.

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

cpLX>

04.27.07 ELECTRONIC DESIGN

BOB PEASE | Contributing Editor

rap@galaxy.nsc.com

ED ONLINE 15416

Bob's Mailbox

ello Bob: I’m designing wideband photodiode amplifiers
and using Jerald Graeme’s excellent book (Photodiode
Amplifiers—Op Amp Solutions) as a reference. In the case
of a composite transimpedance amplifier (TIA) discussed
in Chapter 6, do you know whether the phase compensation
requirement that dictates a value of C F for stability/gain peaking
(i.e., the formula on page 58) changes? (Refer to my column
'What's All This Transimpedance Amplifier Stuff, Anyhow?"
(www.electronicdesign.com, ED Online 4346) Be sure to write down
these questions: What BW (min and max) do you need? What noise
do you need in that BW? What is your Z source (R, C)? What is the
minimum and maximum signal size? Data? These questions can
help you define your circuit and amplifier needs. These are usually
defined by real circuits and not by a formula. If one circuit doesn't
solve your problem, you may need another circuit. No book makes it
easy!/rap) I’ve designed and built TIAs before, but not high frequen¬
cy. I’m getting ready to design a photodetector for an analog applica¬
tion to measure pulses (amplitudes) as small as 10 pA and as fast
as 50-ns pulse width (10-ns rise/fall times). (If you have signals as
fast as that, you are definitely interested in op amps with low V
noise, and the I noise probably won't be so important, unless you
get silly. You have to get low V noise/(Z s ) at the high frequency of
interest. And you haven't mentioned your C s , so / can't guess what
yourZ s is. It had better be a small C s .../rap) Right now I have a
spreadsheet (People who use spreadsheets expect some kind of
quasi-linear problem. These TIA problems force you to change your
whole circuit; so it's not very linear, /rap) where the fixed parame¬
ters like the op-amp parameters and diode parameters are entered.
Then R F , damping, and signal current (pulse amplitude) can be
played with. What gets calculated is the C F (compensation), band¬
width, noise (diode shot, amp current, Johnson, amp voltage, and
SNR). The tricky part is the amplifier voltage noise, which experi¬
ences noise gain that dominates wideband designs. That is why the
datasheet for Tl’s 0PA656 FET amp recommends using the
0PA846 (replaced 0PA686) and 0PA847 (replaced 0PA687) bipolar
amps, which have lower voltage noise and input capacitance.
Besides the TIA topology, I entered in the composite topology and
bootstrap but didn’t see a real advantage to those. The composite
reduces noise bandwidth (and signal bandwidth) with a second op
amp in the loop with the TIA, so it probably has some advantages
over a separate filter after the TIA. (I have never been enthusiastic
about that approach— not a winner, /rap) The bootstrap just made
the pole from the compensation capacitance (which is a sum of a
few capacitances) dominate versus the second-order pole from the
limited open-loop gain and feedback zero. This second-order pole is
the bandwidth given by the standard TIA bandwidth formula with
45° of phase margin. But if the R F , C F pole is less than that, it domi¬
nates, and bandwidth corresponds to 1/(2*tt*R f *C f ).

• John Dailing

• Pease: Let me know if you have a problem. Sometimes,
some foolish person sets you a task that cannot be accom¬
plished. Then you need a friend to explain "this has become
impossible..."

Hi Bob: When I read your March 1 column ('What's All This 'Oth¬
ers Stay Lighted' Stuff , Anyhow?" p. 20), I had to e-mail you about
phasing generators to the grid. With all the wind generators being
installed at remote locations, phasing would be a large problem,
except wind generators use induction generators. An induction
generator is just a three-phase motor (or single-phase motor) that
is run above synchronous rpm. Since an induction generator has
to run 5% to 8% above synchronous rpm, there is no phasing
needed. Just get it close to synchronous rpm and connect to the
grid. If the wind generator rpm is low, it will act as a motor and
bring it up to speed or load it down to the correct rpm. No phasing
or governor is needed. At high wind speeds, the blades stall so
the generator does not over-rev. At low wind speeds, the generator
becomes a motor and keeps the speed up.

• Steven Schmitt

• Pease: You're saying that if an induction generator's synchro¬
nous speed is 120 rpm (for example) and you bring the gener¬
ator up to 119 or 122 and throw the switch to connect it 20
times, there will never be a huge surge of current? Not
enough to blow breakers? Many big wind generators now use
electronic switching for best efficiency at all rotor and wind
speeds. So, any such system would have its own needs for
synchronization. Your simplified version probably wouldn't
apply. Fair enough? Your statement might apply to small, sim¬
ple wind systems.

Dear Bob: Your recent article about polypropylene caps (What's
All This Capacitor Leakage Stuff, Anyhow?" March 29, p. 20)
raises a question I’ve had for a while about Y5V dielectric ceram¬
ic caps. I got burned a few years ago on a design where I wanted
high capacitance in a low volume and used these near their max¬
imum operating voltage. At that voltage, they only have 10% of
their rated capacitance. So what’s the use of these things?

• Mike Partridge

• Pease: At low voltages and at room temp, i.e., cheap con¬
sumer stuff. They are pretty lousy, but the world has a lot of
applications for a lousy cap. So these do sell some. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

05.10.07 ELECTRONIC DESIGN

BOB PEASE | Contributing Editor

rap@galaxy.nsc.com

ED ONLINE 15498

What's All This Dream Stuff,
Anyhow?

t's 3:50 a.m. And as I've said before, if I can't get back to sleep after lying there help¬
less for 45 minutes, I might as well get up. In this case, I had a dream that was even
worth typing. I have about two Safeway bags full of handwritten dreams, plus 1 20
files of dreams typed. Here's the good, wild dream I had recently.

I was starting down a dark, twisty road after midnight. A fast car
was just ahead of me, and I drove fast to catch up with it. (I actual¬
ly do this, occasionally.) I had some kind of computer screen that
showed curves here and there and then
blank spaces between curves. (I don’t have
one of these.) What kind of cars? Unde¬
fined—it was dark.

After a while, I sorta caught up with the
guy, who let me pass and hollered, “Great
road!” And I hollered “Yeah!” I kept driving
fast, and after 10 minutes, I got a little
ahead of him. I looked out across the dark
valley. I couldn’t see if it was 10 miles
long or 40.

After a while I stopped driving, went in
a house, and looked out a big picture
window—8 feet high by 30 feet wide. I
could see it was only a few miles of val¬
ley, but I couldn’t see if it was farming or
ranching. Not many lights. Dark night.

I descended to the right-hand corner
of the picture window in the dark and flipped the
catch, slid the window open, and climbed out to a lower level. (I nev¬
er do this in the real world.) The guy who had been in the following
car followed me down through the room and window in the dark.

I walked stealthily through the second living room and opened
the second big sliding window. And again, and again, for about
eight windows. Then at the last one, I quit.

In the morning, I looked out to see a very big dam, right beside
us. It dropped off over 120 feet and was over 500 feet wide, and
there was a lot of blue water going over it. The other guy was
sleeping in the passenger seat of a car beside the water. A couple
of talkative women came over in kayaks and tried to chat him up.
He said nothing. He motioned with his finger, “keep quiet,” and I
said nothing and kept my head down.

After a while, I drove up through town. It seemed to be a small
Colorado town at the foot of a big (dammed) lake. I drove through
town a couple times. I saw a sign that said “Witch .... Cup.” The
second time, I came to a slow intersection where long trucks were
waiting to turn left. I wedged in on their left and made a left turn.
Then I could read the sign—“Witches’ Brew Cup.” What the heck is
that supposed to mean?

I came down to the foot of the town and saw a huge sailboat
with a wide extruded “mast” 100 feet tall. At the top and the bot¬
tom were rollers, to roll up the “sails,” which unrolled like Venetian
blinds. I understood immediately that there was a big cup, race,
and regatta (sponsored by this local
“Witches’ Brew”) with 10 identical
boats. Coming up soon.

As I approached the bottom of
town, I looked up at this sailboat on
the left and prepared to turn right. I
used a fancy flat “mouse” to steer the
car to the right. Soon I saw a nice little
visitors plaza with many signs that read
“Witches’ Brew Cup” and “Free Road
Map.” I pulled in to find out where I was.
Of course, as soon as I opened the map, I
woke up. (I still wonder where I was!)

OLORFUL HEADLIGHTS * I some¬
times dream in color, 2% of the time, but
most of this dream was sorta black and
white. Even the “blue” water didn’t really
have much color. It was just dark. Color at
night is minimal in my dreams, though I once saw cars with red,
blue, and green headlights on a highway. When they went behind
a telephone pole, the color changed!

What’s it all mean? Not much, I don’t think, but I write them
down anyhow. Just yesterday I dreamed I was trying to get advice
from three people to find Helen Senk’s house, and they were all
wrong. The previous night I was buying a hazmat suit with a big
zipper, and once inside, I wrote a check for $85.

And what was the earworm of the night? As I said back in ’99,
“earworms” are songs that often worm their way into your head
and can’t get out. Tonight, Miss Peggy Lee’s “Willow, Weep For
Me” is such a pleasant song. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

05.24.07 ELECTRONIC DESIGN

••

ED ONLINE 15653

BOB PEASE | Contributing Editor
rap@galaxy.nsc.com

Bob's

Mailbox

i Bob: Just read Bob’s Mailbox from the April 12 issue
with great interest—in particular, the letter from Terry
Hosking about audio capacitors and your reply. It
seems to me that the audio field above all others is
one beset by pseudoscience and pure wooly thinking. (I tend to
agree, /rap) As someone who earns his living as an electronics
engineer, but also was a classically trained musician in the dim
and distant past, I am very much aware of just how subjective
any judgement of audio “quality” is likely to be. In particular, I
cringe inwardly when I read or hear audiophiles raving over this
amplifier’s total harmonic distortion (THD) figure being 0.001%
lower than that amplifier’s figure. It may well be, when measured
under certain conditions. But—and this is a big but—in a com¬
pletely blind test, would anyone be able to consistently judge one
as better than the other? I somehow doubt it. (I have actually
seen cases where distortion as small as that can be heard and
appreciated. If you look at the LM4702 data sheet, there is a
basic audio amplifier ; where the negative input of the LM4702 is
coupled to ground through 20 p F and 1.8k. If you put 200 pF
across the 20 p F, it sounds a lot better because electrolytics can
generate distortion. Now, I measured the distortion at 20 Hz
caused by the electrolytic, and it's about 2 ppm. Yet I have heard
people with good ears say that the music sounds (slightly) notice¬
ably better if the 200 p Fis bypassed with a wire shunt. (Also, I
want to do some tests with 2000 pF, but everybody's too busy to
listen.) I am sure the distortion they hear is not the 2 ppm of dis¬
tortion I measured, so it must be something else. But still, it is
probably very small. The LM4702 has a THD of about 3 ppm at
its output. Some people say it sounds as good as any power
amplifier they have ever heard. That may well be so. It sure is
clean with low distortion, with a clean transient response./rap)
What I would be interested to know, and maybe you or your read¬
ership may be able to help with it, is whether any such tests have
ever been carried out in proper, controlled circumstances. And if
so, what were the results?

• Chris Hill

• Pease: I have tried to set up some controlled A-B tests, but
most people seem to be too busy to do such tests. Thanks
for the comments.

Hi Bob: Can you suggest any reasonably comprehensive
sources of information on the analog characteristics of “pas¬
sive” components—for example, the dielectric absorption of
surface-mount capacitors? I recently ran into a problem where
a supplier of assembled printed-circuit boards (PCBs) substitut¬
ed a polyester cap for an obsolete polycarbonate in an integra¬
tor circuit. Changing the PCB layout was not an option, so the
cap was predefined as 0.1 pF in a 1218 package. The available
parts were Mylar (polyester—cheap but bad), PEN (polyethylene

napthalate—better), and PPS (polyphenylene sulfide—good
enough). (My experience is consistent with yours. PPS is the
least bad of all the ones you mentioned, as wave-solderable. The
polypropylenes are much better, but not so small, and not wave-
solderable./rap) My setup was cheap and dirty: ±5-V power
supply/single-pole double-throw (SPDT), center off switch/lk
resistor/cap under test/LPC661 buffer/DMM. The procedure
was to switch the device under test to +5 for 1 minute, to 0 for
1 second, then open, and watch the DMM. (That is basically the
right test, but you might need to change the times if you were
using the cap in a different timeframe, /rap) I also learned that
all of the surface-mount film caps that I tested were useless
after reflow soldering, unless they were immediately and prop¬
erly cleaned with hot Dl water.

• Ben Barnett

• Pease: Huh. I guess I didn't know that.

Bob: I assume from your comments in “What’s All This Cold Toes
Stuff, Anyhow?” (April 27, p. 20) that you have diabetes. (Correct,
/rap) If so, you might want to think about changing doctors. Periph¬
eral neuropathy is indeed a common complication of diabetes, com¬
mon enough that all diabetics should be warned about it and what
consequences it might have, even if they don’t yet have it. (I guess I
should ask for a good contact fora diabetes specialist. Since I was
diagnosed with diabetes, I have hiked for more than 12 years, and
in cold weather, and in insane conditions, for thousands of miles
and never had any trouble until January, /rap) Furthermore, it’s pos¬
sible (even easy) to test for peripheral neuropathy during a regular
checkup. This should be done, and any patient who develops signs
of peripheral neuropathy should be warned again, more specifically.
(Uh, yeah! I will check into this. Note: hiking hard and burning off
the sugar (and taking in extra sugar) has given me some of the best
health of my life. So I have still done a lot of things right./rap) What
happened to you shouldn’t have happened, unless you were
warned and ignored those warnings.

• Fred Webb

• Pease: I was not warned. But what happened to me was a
gradual deal. How the hell do you give a frog a warning about
warmer and warmer water? Just today, the pain went out of my
left foot. Don't ask me to explain it. The right foot lost the pains
a week ago. So while I am still angry at myself, and only myself,
I am not as mad as I was. Thanks for the advice. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

06.07.07 ELECTRONIC DESIGN

BOB PEASE | Contributing Editor
rap@galaxy.nsc.com

What's All This
V BE Stuff, Anyhow?
(Part 2)

've been debating with a guy who argues that a transistor won't work as a transistor
unless its Vce is bigger than its Vbe. He keeps reading this in books. Also, he points
out that if the base and collector are nominally tied together to make a diode, you
might think that it's okay. But actually, he says, the I x R in the collector path makes the

V CE lower than the base voltage, so it
won’t work. Well, I’ve been looking in
some of those books, and they some¬
times do say that. But when they do,
they’re wrong.

When a transistor’s V CE is slightly
less than its V BE , it keeps right on
working like a transistor. Can I prove
this? Sure. Look in the NSC linear
Databook at circuits such as the
LM10. The LM10 wouldn’t work on a
1.1-V power supply, if the transistors
aren’t working well with V CE as low as
350 or 250 or even 150 mV, which is far below V BE . Of course, you
have to be a good engineer to make these circuits work well.

NSC guys (like Bob Widlar) have been doingthisfor40 years.
Look at the V CE curves of any transistor. When V CE falls below V BE ,
it’s not a disaster. Put a transistor on a curve-tracer. Apply a bias
like 1 pA per step to the base. When you change the V CE from +1.0
V to 0.6 V to 0.5 or 0.4 V, l c doesn’t change much, does it?

Okay, maybe when you get V CE down to 0.35 V, the gain starts to
degrade some. But above that, at room temperature, it’s not a big
deal. There is no demarcation between V CE > V BE and V CE < V BE . No
inflection. The beta doesn’t even change more than perhaps 2%
per volt, and it does so smoothly.

Now run the temperature up to 125°C. Can you design a cir¬
cuit that works up there? It’s not easy. But if you don’t need a lot
of swing, some specialized circuits work just fine. Look at the
LM4041-1.2 orthe LM185. Many of their V CE s are about 0.3 V,
yet they work hot and cold.

How about 160°C? How about 260°C? I can’t, but Widlar
could, and did, in the LM12. After all, in the old days, a pentode
could run with a very low V PLATE —much lower than V SCREEN . It’s
hard to comprehend this, but after a while, you get to understand
and believe it. It’s an analogous situation that the output voltage
is so low, you can’t believe it will work. But it’s true. It does work.

NEXT TOPIC • When V BE = approximately zero, changes in V BE
certainly have no effect on l c , right? Wrong. In Widlar’s LM12,
some of the transistors are so biased, when V BE = 0, the V BE can
still influence the collector current (and vice versa) whether the
V BE is a few millivolts positive or negative.

Admittedly, you can’t see this easi¬
ly in a silicon transistor at room tem¬
perature. But you can see this in a sil¬
icon transistor at 220°C, or in a
germanium transistor at room tem¬
perature, which is about the same
idea. Go ahead and measure it.

When I did, I was impressed by Bob
Widlar’s brilliance.

Also, the beta of a transistor can
still be important, even when V BE is
about zero. That’s because as V BE
moves up and down a few millivolts
compared to zero, the base current needed may be small, but
finite—not negligible. The base current and its changes are neces¬
sary. And if you start at l B = 0 and pull the base negative, the col¬
lector current can decrease.

I must remind you that high-beta transistors (300 and up) still
have disadvantages in terms of voltage gain or mu. When the beta
gets too high, and because mu is inversely proportional to beta,
the voltage gain is hurt. I remember a test that asked how much
voltage gain a particular amplifier design has. The answer was
supposed to be 20,000. But the gain was really 9000, as the
betas were too high and the Early Effect was too strong. I passed
the test after I explained my solution.

As a rule of thumb, I use mu X beta = 2 million. On some
devices, that product is only as good as 1 million, or even 4 mil¬
lion on LM194. If the beta gets better, the mu = l/h RB gets
worse, and the voltage gain suffers. Be careful not to allow in
transistors with too high beta in circuits where poor mu could
cause poor performance. Beta is often important. Too much of it
can do harm. So can too little. ©

Part one appeared in Electronic Design's Analog Applications Supplement,
June 26, 2000. Go to www.national.com/rap/Story/vbe.html.

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

The upper curve represents a 2N3904 biased up with l B = 7 pA,
V BE = 700 mV, and l c = 1.05 mA. As V CE decreases from 0.7 V
(at the right-hand arrow) to V CE = 0.25 V (at the left-hand
arrow), this transistor obviously is acting like a transistor in its
active region, even though V CE is well below V BE . Of course!

06.21.07 ELECTRONIC DESIGN

BOB PEASE | Contributing Editor

rap@galaxy.nsc.com

ED ONLINE 15881

Bob's Mailbox

ob: An induction motor pulls nameplate current at nam-
plate rpm. If it is being used as a generator, the slip is
opposite what is experienced when the device is used as
a motor, but rated power is produced at about the same
absolute slip. If connection is made between the mains and
motor/generator at any speed between rated motor full-load rpm
and synchronous speed plus rated slip, the current must be equal
to or less than nameplate current. The service factor rating even
provides a little additional slack for speed matching. There will be
no huge surge of current. There may be issues with the control sys¬
tem if it comes up in the wrong quadrant, or the gear box may not
like the torque reversal under the same conditions. A motor used
for this service should certainly be as efficient as can be afforded
(But that is always true—a tautology, /rap) and that feature will
decrease the full-load slip. Engagement at near synchronous
speed is pretty easy for an induction machine. I often start the
engines of small induction generators from a dead stop by plug¬
ging the contactor to turn the engine over. Sometimes the engines
do not even have starters installed. (Check. I often start my Beetle
by popping the clutch at a low speed (6 mph in second or 10 mph in
third). I've been told this does less harm/damage/cost than turning
the key and exercising the Bendix. I tend to believe it. But I have only
replaced a Bendix once in 45 years/1,500,000 miles, /rap)
Remember that induction motors are often started across the line
while attached to full load. Lucky motors get reduced voltage, par¬
tial winding, soft-start, or Y delta starters. Most oil-well-pumping
units and gravel crushers just get oversize magnetic contactors.
Starting current is limited by long lines or the service transformers.
(My father had a circular table saw, and I remember when it start¬
ed properly by turning on its switch. But when it refused to start
(bad cap or bad winding?), he just blocked up the motor ; spun the
pulley hit the switch, and lowered the motor down against its pul¬
leys after it started—not a big deal! Just a normal trick! I've learned
how to do it./rap) The control system in most wind turbines is far
less sophisticated than you would expect. (I have been told that
some have synchronous converters—what, 2%, 4%, 8%, 16%?
/rap) Those guys have come a long way, but they have a long way
to go. Don’t give them undeserved credit, and don’t give them my
address. Sorry to hear about the frostbite.

• John Carroll

• Pease: Not a big deal. I'm gaining on it.

Hi Bob: I just read Bob’s Mailbox from June 7, and I must make
a point about total harmonic distortion (THD) in audio systems
that many people miss. What really matters are which harmon¬
ics and non-harmonically related overtones are generated dur¬
ing distortion. For example, crossover distortion (caused by an
underbiased class AB amp) is much more objectionable than
clipping. That is to say that if the distortion is caused only by

clipping, a higher percentage of THD is permissible than if the
distortion is caused by underbiasing. (We agree, of course, that
some kinds of distortion are more important than others. But
when the distortion gets down to 3 or 1 ppm (not to mention 0.3
ppm), I don't think the distortion is important anymore. Take a
look at the datasheets on the LM4562 and LM4702./rap) One
attempt to weight the harmonics to make a distortion measure¬
ment scheme that reflects subjective evaluation is the GedLee
metric (www.gedlee.com/distortion_perception.htm). From
what I understand, it correlates very well with subjective ratings.

• Stephen L. Martin

• Pease: Thanks. Best regards.

Hi Bob: I just got back from a 500-mile hiking trip on the
Appalachian Trail and saw your piece on the cold-toe detector
(“What's All This Cold Toes Stuff, Anyhow?" April 27, p. 20). I
would think this would be an ideal application for a small micro¬
power microprocessor. (No, it would not be “ideal," because I do
not do microprocessors. I do human interface./rap) Using one of
these, the parts count would be reduced to four: the temperature
sensor (My scheme can accommodate four or more temperature
sensors./rap), processor, signaling device, and small button cell
for power. It could be made small enough to mount on the shoe,
say, where the tongue of the boot starts. (When the tongue of the
boot is covered with a foot of snow, I'm not going to peek to see if
the LED is blinking. I want it to beep in my ear. Besides, my boot is
usually covered up with leggings./rap) (I’m not too keen on the
idea of running wires up my legs!) A piezo speaker or flashing LED
could warn you the temperature at the toes is below freezing. (No
way! I want to know when the temperature is dropping! Your men¬
tality that a computer can tell me that my feet are already too cold
is exactly what's wrong with all the digital guys. Sorry, but go away.
Get thee behind me, Satan!/rap) The advantage of this approach
is the processor can be programmed to take a temperature read¬
ing at given intervals and then to take readings at shorter inter¬
vals if the temperature approaches a dangerous level. When not
active, the circuit would draw microamps, making a small button
cell practical. The only downside of this approach is the need to
write a simple program and programming the processor with it.

• Steve Weber

• Pease: I've never made any mistakes writing software. Because
I never write software. And I'm not going to start now. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

07.05.07 ELECTRONIC DESIGN

ED ONLINE 15963

BOB PEASE | Contributing Editor
rap@galaxy.nsc.com

What's All This Production
Costs Stuff, Anyhow?

A guy asked me how to optimize a bridge circuit. His design had eight op amps
and over a dozen precision resistors. I showed him howto minimize the number
of (expensive) resistors and precision op amps. I got it down from eight to four to
two. Then I even figured how to get the number of op amps down to one.

But before we finished, I had to inquire
how many bridge circuits he was planning
to make. That’s important. If you’re only
going to make five or 10 or 20 of them, you
will probably want to just get the circuit built
and running and evaluate the system you
are going to run. The cost per unit is not
important. The time you waste optimizing it
could be very important.

If you’re planning to make thousands of
them, you’ll want to put in some extra engineering effort to fig¬
ure out how to optimize and minimize the cost of each compo¬
nent—which usually isn’t important if you’re going to build only
a few. The cost of parts, the assembly labor, and the trim-and-
test work will be quite important for a large-volume design.

So when a guy asks for help, I have to explain this problem.
Do engineers understand this? The good, experienced ones
do. The new ones (the ones who have to ask a lot of questions)
have to learn. Often, trying to be very frugal and using the
cheapest parts is poor economy. Spending more for some pre¬
cision parts may save you a lot of grief.

It’s very much the same as a guy asking, “How can I get the
best low-noise amplifier?” As I always respond, “What’s the
bandwidth, what’s the source impedance, and what’s the size
of the smallest and largest signals?” Without that information,
you can’t optimize anything. So the planning is really important.

PLAY BY THE RULES-IF YOU KNOW THEM • When I
was back at Philbrick, we had several plans for optimizing a
potted module. One plan was to optimize parts cost plus 150%
of assembly labor at 2 cents per second. Back in 1970. But
one day, I was told my new circuit was badly designed because
I was violating the new rules. “What new rules?” I asked.

They forgot to tell me that they had changed the rules. It was
now 150% of parts plus 450% of labor and assembly costs—
and they hadn’t bothered to tell me the rules had changed. The
word infuriated isn’t strong enough to describe my mood. How
the heck am I supposed to do my job with poor information?

Another time, they changed the “rules” so that the cost of a
jumper went up (or down), while the cost of a double-sided pc
board went down (or up). But they didn’t tell the engineers. So I
confronted our manager, Richard. “These new rules mean I

should avoid a double-sided pc board and put in a
few jumper wires, right?” After he thought about it
a second, he saw red, and said, “No, that would be
wrong. Let me get that fixed.”

Well, in a couple of weeks, Richard was gone,
and the question wasn’t solved. And a month
later, I was gone. I walked out on the last day of
1976. If you’re hired to optimize new designs,
but the rules keep changing—and they forget to
tell you the rules—hey, you have to walk out.

PLAY BY THE RULES, PART 2 • My friend Arnold had
designed a very good high-voltage amplifier with ±100-V out¬
put swing, the Teledyne Philbrick 1022. But the guys in mar¬
keting decided they needed a low-price version to fill a high-
volume need.

So Arnie took all the rules and figured out how to use a more
spacious layout (so things weren’t packed in so tight) with low¬
er assembly costs and lower-cost parts to make the Model
1032, with definite cost improvements.

But after it was put into production, the Manufacturing
Department decided to re-interpret the rules, and the Model
1032 was listed as more expensive than the 1022. Arnie could¬
n’t win. He bailed out, too.

BOOK REVIEW • Check out Wideband Amplifiers by Peter
Staric and Erik Margan. Available from Springer.com for about
$159, this 612-page book covers all aspects of high-frequency
amplifier design. It is now in its second edition after correcting
a few typo errors. The first printing sold out. I like its attitude
and its insights. It covers theoretical and practical applications,
computer-aided design and filters, real semiconductors, and
integrated circuits. It also covers bipolar transistors, FETs, fast
op amps, and current-feedback amplifiers. I’ll recommend it for
anybody working in this field. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

07.19.07 ELECTRONIC DESIGN

••I

BOB PEASE | Contributing Editor

rap@galaxy.nsc.com

Bob's Mailbox

ob: Several years ago, an acquaintance bought a very
expensive “high-end” audio system. The speakers were
huge. The amplifiers were 350-W “mono blocks” using
eight 6550s in a push-pull parallel configuration. The
wires were about $100 per foot. Within a minute of first operating
the system, the amplifiers were releasing smoke. He called the
amplifier manufacturer. Their first question was “What cable are
you using?” The strange impedance of the fancy cables interacted
with the feedback system in the amplifier and caused an ultrason¬
ic oscillation that destroyed them. So maybe the cable makes a dif¬
ference, not because of its characteristics but its action on the
amplifier and speakers. Not many amplifiers have a truly zero out¬
put impedance, especially vacuum-tube types.

• Bob Bodmer

• Pease: I have heard people say that low-impedance speaker
cables have a lot of capacitance per unit length, if you just
measure the cable open-circuit. “So when you run an ampli¬
fier that doesn't like capacitive loads, it can oscillate and
cause great damage ," these people say. First of all, vacuum
tubes are supposed to be able to tolerate overloads in any
good design. Second of all, a low-impedance speaker cable
does not look like a capacitor unless you run it with no
speaker. My favorite speaker cable is to parallel 20 strands
of wire, going out to the speaker, and each wire serves as a
twisted pair with one of the 20 (paralleled) return wires that
come back to ground. If you run this with no 8-Q speaker, it
will indeed act capacitive, and the amplifier might be
unhappy driving the many thousands of pf. It will look like
an unterminated transmission line, and the reflections
could be nasty. But when connected to an 8-Q speaker, it
looks like 7.5 Q because it has the characteristics of a 7-0.
transmission line. So every amplifier should drive this cheer¬
fully. The cable and load will act like 7 or 8 Oat all frequen¬
cies, unlike lamp cord that acts like an 80-0 transmission
line at high frequencies. Audio amplifiers are not required to
have a 0-0 output impedance. Any low impedance that is
consistent and predictable can work just fine. An audio
amplifier that wouldn't drive an 8-0 load sounds pretty flaky.
And if a customer complains and the amplifier maker tells
him “ Oh, we forgot to tell you, our amplifiers don't like cer¬
tain kinds of loads," but they didn't put it in their user's
manual, that is a poor way to do business.

Bob: I saw your comments about Y5V dielectric in the May 10
issue. I have always avoided it, not only for its -80% voltage coeffi¬
cient but also its poor temperature coefficient. For general-pur¬
pose use with 20% tolerance, I have always specified X7R dielec¬
tric or its more recent derivative X5R. However, recent cost-saving
process technology changes mean you have to be wary. Time was

when X7R capacitance was -20% at rated voltage. Now with mod¬
ern thin-layer printing technology, the change of capacitance is
-50%. So be careful. If your design needs 20% tolerance, you
need to double the rated voltage of the X7R capacitors you use.
For a 5-V supply, use a 10-V part, not 6.3 V. It is better to do this
than to double the capacitance. The cost is about the same, and
so is the capacitance at working voltage. But the reduced coeffi¬
cient of the higher-voltage part makes for a more stable design.

• Chris Green

• Pease: Yeah, I believe it. Many “ improved" processes are
worse than ever. Your solution seems reasonable.

Dear Bob: I just wanted to write you a quick note commending
you on your “What’s All This Capacitor Leakage Stuff, Anyhow?”
article (March 29, 2007). From the title I suspected that the
devil of the measurement would be in the (attention to) details,
and it was! When I read the sentence “I also keep my left hand
grounded to +8.8 V,” I was thoroughly pleased! You think like an
electron, and that is a good thing.

• Chuck Woringer

• Pease: I tried grounding my left hand to +1 V. It made tiny
errors, but when I did it nine times in a row, it made very small
errors, and I didn't even want to make tiny errors. I thought
about finding a teflon-insulated ball-hook or teflon-insulated
alligator clip, and that ain't easy. By “ grounding" my left hand
to +8.8 V ; the problem shrank to an acceptable error.

Dear Mr. Pease: One of your columns (Jan. 18, 2007) com¬
mented that it will be a longtime before you will change to SI
units. You are in good company. A few months ago, I was curious
about the pace of conversion to SI. I noticed an important
organization that tried metric units, and now it changed back to
U.S. Customary units. The document is called “Caltrans Metric
to U.S. Customary Units Transition Plan” (www.dot.ca.gov/hq/
oppd/metric/TransitionPlan/Transition-Plan-021105.pdf).

• Jim Marks

• Pease: Yeah, I peeked at that 29-page document, and sent it
to print out, and will read it later. The introduction says it will
be less trouble to go back to “all U.S. units" and less likely to
lead to the delays or extra expenses that happened in metric
conversion—and safer, as of the era 2005-2007. Maybe we
should share this philosophy with the Brits and Aussies. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor CorpSanta Clara, California.

08.02.07 ELECTRONIC DESIGN

ED ONLINE 16201

BOB PEASE | Contributing Editor
rap@galaxy.nsc.com

What's All This Canyon Turn
Stuff, Anyhow?

et's say you're flying a light plane, and you fly up into a canyon at 8000 feet. At
8000 feet, your turning radius is perhaps 600 feet. The canyon is 1 000 feet
wide. You'd like to turn and qet out of the canyon, but there's no room to turn. And
the canyon floor is rising too fast for you to climb out. Are you dead?

Maybe not. I have been studying this for sev¬
eral years. Our good friends lost their daughter
in a flying accident in a box canyon, just like
that, about 10 years ago. She and another ♦
student pilot died in the crash, along with *
the certified flight instructor. I still cry about f
that. A lot. Especially today. How can you
get out of this trap?

(a) Put down your flaps to 10° or 15° (or maybe
20°) to increase lift and drag. Apply full throttle.

Do not let speed build up. Do not let your alti¬
tude above the canyon floor get small. Turning
radius may be 600 feet at full speed, but it
may be smaller under those conditions. You
can turn a lot better. Flaps and power help you maintain control
and avoid stalling. Maintaining some altitude above the ground
can be extremely important in view of (c).
b) Get way over to the right of the canyon, maybe 40 feet from the
wall. I say to the right, because the flight instructor (or pilot in
charge) on the left should be in charge of these fairly danger¬
ous maneuvers. He has to use his best judgement. His feel for
what the plane is doing is very important.

When he cocks the ailerons and pulls back on the yoke, it
may be very close to a stall. You gotta have a feel for that. It’s
also true that the engine’s torque may help you turn left bet¬
ter than right. (If the canyon is deepest on the right, okay, get
over to the left. Or if you know there is a crosswind from the
right, turn toward it.)

(c) Make your turn early, make a hard turn, and let the plane
descend. Use the plane’s lift to pull you around in the tightest
possible safe turn that you can do without stalling. If you tried
to hold altitude, you will probably stall and crash. But if you let
the plane descend to the left, you have a better chance. That’s
assuming you are somewhat above the canyon floor.

(d) Do not fly up into canyons. Never. If you want to check out a
canyon, always come in from above and descend—assuming
the canyon does not have excessive sharp turns or narrows (of
which there may be many).

(e) Under conditions of full power but lowered flaps, you may be
flying at 55 or 45 mph depending on the headwinds. So even if
planes aren’t exactly as bust-proof as cars, you have some
chances of surviving a 50-mph crash. Not so at 85 mph.

The flaps may get you around the corner. There’s no guaran¬
tee, but that’s what I would bet on. Better odds—and a
better chance for a good pilot’s skill to help.
Of course, a really good pilot would never
get you in that bind in the first place. A cat
will not crawl into a hole where his
whiskers scratch the walls.

(f) If you live in flat country, you may never
have to worry about this. But if you live near
mountains, you would definitely want to practice
this at a safe altitude, out in the open, so if you stall, you
have a lot of air under you to recover. You may not be able
to judge how much space this maneuver takes, but you can
get a feel for how the plane feels and handles.

(g) If you were in a very narrow canyon, maybe 200 feet wide, you
could zoom up, cut speed, apply some flaps, do a half-roll, cut
power way back, and finish your half-loop. But this is not legal
for most light planes, and it might cause excessive G’s or
speed. It still might save your life—if you were 600 feet or more
above the canyon floor.

MOUNTAIN PASSES • Do not fly directly toward a high
mountain pass trying to get across. If you have a large altitude
margin, you might do that. But if you aren’t sure, don’t try to fly
toward the pass.

Circle around on the right. If your altitude is okay, then you
could fly up to the pass and go ahead and veer right across the
pass. But if you are kind of skeptical, and you think there may be
downdrafts, or if you aren’t sure there will be enough updraft, veer
left and go around again. Try to get a little higher and burn off a
couple gallons. You may have much better luck on the next try.

Too many pilots have failed and crashed because they were con¬
fident they could get over a pass. But downdrafts can be very nasty
and can be the downfall of “confident” pilots. The Alaskan bush
pilots who are alive today have learned from such advice. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

08.16.07 ELECTRONIC DESIGN

BOB PEASE | Contributing Editor
rap@galaxynsc.com

What's All This
Anyhow?

ED ONLINE 16566

God Stuff,

here's an old story about an American visiting Ireland. As
he was imbibing a beer at a tavern, one of the locals
asked him, "Now, are ye a Catholic or a Protestant?"
The American replied astutely, "Neither. I'm an atheist."
the canny Irishman asked sharply, "Ah, but are you a

Then

Catholic atheist or a Protestant atheist?

These days, trying to prove whether
you’re a Sunni atheist or a Shiite atheist
in Iraq doesn’t sound very easy, either.
Yet it might be important, depending on
who’s asking the question.

I am not going to argue with you about
your religion, or any version thereof, or
any lack thereof. Whatever you like to
believe in is fine with me. As near as I
can tell, there is a very wide distribution
of religious belief within the engineering
and scientific community. Also fine by
me. I’ve heard some people argue that if
you believe in evolution, you can’t
believe in a religion—and vice versa. I
don’t agree with that correlation at all,
and a lot of people don’t.

Do I believe in God? Yes. I am a Christ¬
ian and a member of an Episcopal
church. You may have heard that in the
1860s, every Christian church in the
U.S. split asunder over the issue of slav¬
ery—except for the Episcopalians. Appar¬
ently, they thought they could “get
along” despite serious differences.

“God has been good to me, Alleluia”
(do re mi mi mi mi, re, do re re)

“Let me return the favor, Alleluia”

(do, re, mi mi, re, do, re, do)

• I’m not

going to argue with people who say they
are atheists, or agnostics, or any particu¬
lar religion. But I am in favor of God. “I
will try to help God,” and I think God will
encourage me to do the right thing.

What is the “right” thing? Everybody
has his or her own moral compass. I don’t
want to argue much about this. Exactly
where your moral compass is, or where
you got it from, is your business. So long
as it works, that’s fine with me.

But eventually you might ask yourself
where you got your ethics. Did you get
them from a church or a Sunday school?
Did you get them from your mother or
father? Whatever way you got them,
thought about them, and refined them is
fine by me. Even reading Dilbert can bring
you to conclude that some of the charac¬
ters in that comic strip learned their ethics from some strange places. Learning to not
do what the pointy-haired boss does is a pretty fair way to learn ethics.

When we ran the 1909 Rutherford experiment in our 1960 physics lab, we learned
that bombarding the nucleus of a gold atom could lead to some knowledge of the
structure of the nucleus. You bombard a beam of alpha particles off a gold atom’s
nucleus and see at what angle they bounce back. The distribution is quite educational.

Similarly, when problems are
bounced off of us, we don’t have to
write down how we define our ethics.
But eventually, by circumstances, we
show what our ethics are.

• Here at NSC, we
have an intranet course on business
ethics that is fairly good. We had to
study certain intercompany relation¬
ships and figure out how to be fair to
our customers. All employees are sup¬
posed to take its test until they pass.

It’s fairly educational. I don’t think Wally could pass it, nor the pointy-haired boss.

Do you believe in (most of) the 10 Commandments? That’s good for a lot of real-world
cases. How about the Golden Rule? Many of us agree on that—most of the time. How
about the IEEE Code of Ethics (www.ieee.org/portal/pages/about/whatis/code.html)?

I tried to find help there a couple of times, but didn’t find much.

As long as your moral compass works, that’s fine with me. Most engineers (and
most people) have figured out that being nice and fair to your customer is a good
idea. “Screw the customer” has long been recognized as a poor business practice.

So, I won’t try to argue with any reader about religion. But I tend to be in favor of reli¬
gion. And its positive side. I wish we could all avoid its negative side. In many places,
Protestants and Catholics have learned to get along. Even in Ireland. “Love your neigh¬
bor” is a nice theory, but stopping “hate your neighbor” may be even more important.

What religion (if any) do you believe in? It sure is none of my business. But it would
be nice if your religion allowed you the same respect for my beliefs as I have for yours.
I don’t denigrate your religion. If the U.N. Charter promises freedom of religion, does
that allow your religion to denigrate mine? I’d hope not.

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.0. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at National Semiconductor
Corp., Santa Clara, Calif.

09.13.07 ELECTRONIC DESIGN

ED ONLINE 16780

BOB PEASE | Contributing Editor

rap@galaxy.nsc.com

Bob's Mailbox

M any people wrote in about
“What’s All This Canyon
Turn Stuff, Anyhow?” (Aug.

16, 2007, p. 16.) Some sug¬
gested the pilot should do a Chandelle,
an Immelman, a Wingover, or a Ham¬
merhead Stall. Tony Schrock suggest¬
ed www.scootworks.com/rdrc/aero-
bat.html for illustrations of these and
other maneuvers. But as the plane was fully loaded and underpow¬
ered, and the elevation was 8000 ft, and the day was hot, none of
those were going to work.

Some people said my 20° flaps setting was about right. Others
said that full flaps might even be better. Some guys said, “Line up
above a road when you practice this over low country, and see how
far offset you are after your turn.” One guy suggested using a GPS,
and that should work even better.

Everybody agreed that you never want to fly up into a canyon,
and even flying down through a canyon could be very dangerous
unless you were positive there were no bad kinks. One guy, howev¬
er, said that my theory that a cat won’t crawl into a small hole if its
whiskers were scratching the sides was just an old wives’ tale.

Well, that’s what my mother told me, and maybe she was wrong.
Here’s a related story.

Hello Bob: I read your article on the canyon turn. I too had a good
friend perish in a similar event in the Sierras. He was my boss at
my first job after high school. I then moved on to become a pilot
myself after a short run in the Navy and made a career of it until
starting my current company. His memory stays with me to date. I
had a similar event while on my long cross-country flight for my
commercial license that proves your input here. After a trip to Los
Angeles, Vegas, then Reno, I got up the next morning, and guess
what... snow! I did my preflight checking, weather at destination,
etc. Visibility was visual flight rules, but barely. I decided to load up
and give it a run for the pass. I did go to the pilot briefing room and
studied a local chart posted on the wall. It had the box canyon to
the left of the pass highlighted. My father was an ex-B-17 pilot. I
asked him, being on the right side, to watch for the highway to turn
to the right, marking the pass exit to turn down, to then follow High¬
way 50. Yup, he missed it. I had hit my clock to mark time and
decided that if he hadn’t called it by then, I would head back. Visi¬
bility was lowering quickly, to about half a mile in blowing snow. We
were heading into the jaws of the storm—and the box canyon. I
started my turn and noticed the trees coming into view very quickly.
I ran the flaps now to Max L/D (about 22°) and applied full power
during the turn. The pine cones were now visible. The stall warning
horn was hinting what an idiot I was! I continued the turn (obvious¬
ly, since I am writing you) and made my way back to the airport. We

Nicknamed Thunderbird, this B-17G Flying
Fortress belongs to the Lone Star Flight
Museum, Galveston, Texas, (photo by Master
Sgt. Michael A. Kaplan, U.S. Air Force)

all jumped out with shaking knees
and went for a cold one! Stayed a day
or so longer and made it home safely.

I became a flight instructor and made
it a practice to teach that experience and technique. It works. I
used to fly into canyons practicing the maneuver as far as I could
go and make a turn (in good weather). Thought I would share it with
you. You may save someone’s life or lives. Good for you.

• Michael A.

• Pease: Yeah, man. It's tough to figure out the navigation in
the best conditions, not to mention in snow. I'm glad that
heavy flaps and full power got you around the curve.

Bob: Back in World War II, B-17s flying to England from the U.S.
had to stop for gas at West Bluie Two, a landing strip in Greenland.
Unfortunately, WB2 was many miles up a fjord, and there were
three identical fjords in the area. If you went up the wrong fjord,
you were in a heap of trouble. The only definite way to recognize
the correct one was to fly up one, and if you did not see a sunken
ship 2.6 miles up the fjord, you were in the wrong one and had to
do an immediate canyon turn. Not easy to do in the fog at the end
of a long flight with engines and props configured for maximum
range. Those were real men in those days.

• George Gonzalez

• Pease: Uh, yeah. But at least the plane was lightly loaded.

Was it really that hard fora B-17 with no bomb load to fly from
Newfy to Greenland? Maybe it was. And speaking ofB-17s
again, I saw a B-17 take off from Moffet Field recently , and it
was damn impressive. And not very big. Just tough as nails, as
were the kids who flew them.

FLOATING GAME • As for the test when an oil can was get¬
ting full /Bob's Mailbox/ Sept. 1,2007, p. 18), several people
chided me for using electronics when a simple float would work
much better. But none of them were able to convince me that the
float wouldn’t get jammed or stuck. So, I think one of the electron¬
ic solutions might be the winner. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

09.27.07 ELECTRONIC DESIGN

ED ONLINE 17079

BOB PEASE | Contributing Editor
rap@galaxy.nsc.com

What's All This Safety Margin
Stuff, Anyhow?

ometimes it's easy to tell if you have a safety margin. With a voltage reg¬
ulator, or any linear amplifier, if it was oscillating, you could add a fix —
often, a simple series R-C network from the input or output to ground.
Good. But is it good enough?

To be safe, you should put in a square wave of voltage (or pull
out a square wave of current through a little R-C network) and
make sure that there isn’t any bad ringing. Now, to be quite sure,
you would have to exercise this linear amplifier over its expected
range of voltage and current (and temperature)—and make sure
the ringing stays far away, as I said in Pease’s Principle on page 99
of my book. 1

But how about for a buzzing comparator? If you put in hystere¬
sis, it seems okay. But how can you make sure it is going to run
safely? I don’t think anybody has a solid answer on this. But here
is my solution: change a resistor to cut the hysteresis by about
half. If it still runs safely without any oscillation or screaming as
the input signal passes the threshold, that is a good indication.

So restore the proper amount of hysteresis, and you are proba¬
bly safe. Unlike the amplifier problem, this is unlikely to be affect¬
ed by temperature. But it is likely to be affected by layout, so don’t
let anybody fool around with the layout.

In the real world, those of us who have to drive with snow have
learned that you have to do some practice skidding at the start of
each season and every time there is significant snowfall. As I said
on p. 224-246 of my other book, practice hitting the gas and the
brakes too hard and cornering a little too hard in an empty parking
lot and on the road, too, where snow conditions may be a lot dif¬
ferent. 2 I don’t have to spell that out, unless you are a new arrival
to snowy territory.

JUDGING POWER • Next, how do you make sure you have
enough power-supply bypass capacitance? I have seen a couple
of analytical studies, and they have come to the same conclusion
as my rule-of-thumb solution: Use one ceramic disc cap, 0.02 or
0.1 pF per 1C (on each supply, if it is an op amp with + and - sup¬
plies), and add one 2-or 10-pF electrolytic or tantalum cap per
four or five ICs. But how do you know that it’s safe?

My solution is to lift out (or snip out?) half of the capacitors.
Study some of the critical waveforms before and after you snip and
see if the circuit seems to be okay. Study the amount of ringing on
each power bus. Then, put the caps back in. Of course, some
amplifiers are so slow and docile, they aren’t very dependent on a
lot of bypass caps. But you never know until you check it out.

Why not just leave the capacitors out? Well, you might save a
dime or two. But you would lose your safety factor. You would have
to do a lot more testing at hot and cold to be sure you were safe.

"7T Fen , ase esr is bij een ontkoppelcondensatoi
Figuur4.2 ^^ZmeteenUsgeESRde^dmgsb^
stroompieken trekken, bevordert de
pdcapaciteit met de zelfinducUe van de bus 4

And after your electrolytics have aged, you could lose your safety
factor even at room temperature.

MORE ON FLYING • What if you are approaching a moun¬
tain pass? “If the pass looms smaller and smaller behind the
cowl, you are probably going to make it. But if the pass looms
bigger and bigger, you know you are not going to make it,” one
pilot explained.

That may be literally true, but that doesn’t sound like nearly
enough safety margin for me. Several pilots said they like at least
a 3000- or 5000-ft margin over the pass to allow for downdrafts.

Flere in the U.S., updrafts often go up the west slope of a
mountain, and downdrafts come down the east side—but not
always. And altimeters usually tell the truth, but not always. So
when piloting a plane, you have to have your own rules for deter¬
mining what to use for a safety margin and when to trust it. ©

(1) Troubleshooting Analog Circuits, Elsevier, R. Pease, 1991, p. 99

(2) How to Drive Into ACCIDENTS—and How Not to, Pease Publish¬
ing, 1997, p. 224-246

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California

10.11.07 ELECTRONIC DESIGN

• BOB PEASE | Contributing Editor

rap@galaxy.nsc.com

ED ONLINE 17214

Bob's Mailbox

ob: As we were landing at O’Hare Airport last week, I was
observing the runway and taxiway lighting. I immediately
noticed that the lights are now LEDs. Would you believe
that even these are designed to operate off the standard
6.6-A constant current system? (See www.flightlight.com/airport-
lighting/l.l/1.1.2.html.) This fixture is rated at only 3 W, which
means it apparently drops 0.5 V at 6.6 A (perhaps 1.0 V or more
with transformer losses?). There is still a power saving because
they can supply the current loop from a lower voltage. (Yeah, but
the transformer makes it a game, as you have seen. And they have
to be prepared to put incandescents back in—in case the LEDs get
flaky./rap) There are two ways I could tell these were LEDs. First is
the very pure blue or yellow color. But I can also instantly tell when
a light source such as an LED is operating from rectified ac or
pulsed dc. If you let your eyes scan from left to right, you can see a
crude oscillograph view of the lights, with a dashed-line trail
instead of a continuous trail, indicating a 60- or 120-Hz rectified
waveform. You can also see this effect on LED car tail lights, which
are apparently pulsed to dim them. I estimate the tail lights are
flashing at a few hundred hertz to a kilohertz. (Wouldyou expect
this to be audible? i’d be surprised if it didn't make an annoying
buzz./rap) Someday I would like to build a spinning mirror device
to view this more accurately.

• Ken Lundgren

• Pease: Any simple spinning disc or light chopper could confirm
this. A motor and wheel (black paper) with a notch cut into it
could be much easier than a mirror to rig, ora black-and-clear
plastic foil disc. I bet you could rig up an electric demodulator
too, but not a big deal.

Dear Bob: Seeing Christmas cards for sale in the local bookstore
now gave me pause to realize that the November issues of ED are
being set up. In all my many years behind the desk, only once in
my 35-year career have I worked for a company that actually dis¬
tributed a Christmas bonus—and even then, it wasn’t cash! The
better part of a decade fresh out of school was spent at Litton
Aeroproducts, located near Los Angeles. Litton made inertial navi¬
gation and radio navigation equipment for a vast majority of the
world’s airlines. An occasional customer of Litton Aero was Lot
Polish airline, the state-owned airline of Poland, then under a
loose Communist rule. I don’t recall all the legalities (and perhaps
behind-the-scenes maneuvering) that went on, as I was just a
design wonk. But I do recall that because of some trade
restraints, it was illegal to accept cash from Poland. The average
cost of an inertial nav system at the time was somewhere in the
neighborhood of $85,000, maybe more. What Lot Polish did was
to pay Litton in ham. Yes, the IittIe-piggy-that-couId kind. You could
tell when LP bought something, as perhaps a half-dozen refriger¬

ated trailers would show up in the parking lot—and always, it
seemed around the end of December. And so, with your pay-
check, you would receive a stub in the same envelope that would
grant you one ham. If you didn’t eat ham, you could have the stub
exchanged for a voucher for a 10-lb turkey from the local super¬
market. Litton is gone and never did any profit sharing, but it
makes me wonder what our share would be—ham steak?

• Karl Strauss

• Pease: Hi, Karl. That is funny!

Hi Bob: I had to chuckle when your correspondent first wrote about
filling the oil lamp well (“Bob's Mailbox," Sept. 1, p. 18), since I’d
been mulling over the possibilities of detecting the water level in
my 280-ft drilled well on a real-time basis by either organ-pipe-like
resonance or acoustic time of flight. (Water can come as close to
the surface as 30 ft, according to the label placed on the wellhead
by the driller in 1994. Or to be more precise, 30 ft is listed as the
“static water level.”) (Either of your schemes would work okay , pre¬
suming there is enough accuracy in each range to suit your needs.
My scheme with a capacitance detector between adjacent wires
might work, too. /rap) But considering the original problem, isn’t
the use of a self-heated thermistor a standard, quite low-tech solu¬
tion? (It might work, but it will not give you a gradual indication of
when you are approaching the stop point, will it? Besides, if a little
of the in-pouring oil splashed onto the thermistor, it will cool off,
not so? So it would give a false reading. Would you be able to hide
the thermistor so incoming oil would not hit it? Maybe, but life ain't
as simple as it seems!/rap) Didn’t 100 people suggest this? (No,
thank heavens!/rap) Placing the thermistor in a 555 timer circuit
driving a small loudspeaker would be my approach—no need to
look at an indicator. (Will a 555 drive a thermistor in a self-heating
mode? Maybe so./rap) Digikey has tiny brick-shaped thermistors
with millimetric dimensions and dissipation constants in air of
1 mW/°C for less than a quarter.

• Jim Hayden

• Pease: Maybe your idea would lead to a good solution. Hey, put
two thermistors in parallel (or in series?). The first inflection tells
you to slow down, and the second tells you “Stop!" Go ahead
and build it and send it in as an Idea for Design. If you put them
in a 0.25-in. ID tube, properly vented at top and bottom, they
could detect the rise of fluid and avoid splashes of incoming oil.

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

10 . 25.07 ELECTRONIC DESIGN

BOB PEASE | Contributing Editor
rap@galaxy.nsc.com

What's All This Stability Stuff,
Anyhow?

A guy recently asked me how I would look for a voltage reference that's
stable versus temp cycling. I told him I would take several of the best volt¬
age references I had and use a dual-slope DVM of at least six digits to
compare them to the units in question. He then asked if comparing some

references to some other ones was kind of incestuous. This is
not rocket science.

You take several good voltage references and leave them the
heck alone! Apply some bias and just let them run undisturbed. I
have done this many times. But if possible, measure each of
them, at least once or twice per day. Gather up the trends. Look at
the data. Study them, study the standards, and study the DUTs.

I once had a set of 16 good references, LM399-types, that had
a subsurface (buried) zener and a heater to hold them at 88°C.
Each output was 6.9 V ±2%, and I averaged the output through
some 499-0 resistors. The output impedance was 65 Q. I meas¬
ured the average of eight versus eight. The relative stability
seemed to be very good. The noise seemed to be excellent-
better than 0.1 ppm p-p, out of 7 V, in a bandwidth of about 1 Hz.

But the point is that if you leave something alone, and cycle
something else through experiences, you learn something.
What are you trying to learn? A drift versus temp is easy to spot
if one part goes through a temp cycle and another doesn’t.

Cycle the DUTs twice or four or eight times. Look for trends.
Check the V REF of each of these against the average of the two
or four or eight uncycled references. You may learn something.
What do you see for trends? Do you see a drift that decreases
gradually versus experience? Or is there some hysteretic drift
that keeps coming back?

BACK IN THE DAY • Back 24 years ago, we set up some
excellent long-term stability tests for the LM199AH. We read their
data every week for six weeks, compared to a battery of other ref¬
erence voltages, including ovenized saturated standard cells,

4 X 1.018 V dc. We convinced ourselves that these references
were mostly drifting less than 10 or 15 ppm per 1000 hours.

We sold them with a guarantee of 20 ppm per 1000 hours.
When we ran out of such customers, we shut down that aging/
testing program. But it was challenging. What if one or more of
the tested references had some drift? We were prepared to use
our judgement to decide that the apparent drift of one reference
could be ignored—if all the other references seemed stable.

What if the DVM’s reference started drifting? This was a ratio-
metric test, so we could correct for that. So long as most of the ref¬
erences seemed stable, we could compensate. Of course, we had
to rely on the ratiometric linearity of the DVM, which was guaran¬
teed and inherent, better than 1 ppm. It was a very good HP3456.

To be fair, I shouldn’t just call it a dual-slope DVM. It was more
like quad slope or multi-slope. Its measurement scheme used a
lot of rubbing and polishing. I have seen several other kinds of
six-digit DVMsthat had linearity flaws, but the Hewlett-Packard
ones never showed any such error. Now they are called Agilent.

Since that era, we have used groups of LM399s as refer¬
ences for testing other kinds of band-gap references. Of
course, using a group of four references can increase the
chances that one of the group will start drifting.

But if you have four groups of four, the chances that one will
start drifting and won’t be apprehended are quite small. Long¬
term drift can be fairly dependable. We sent some LM399s to
the NBS/NIST, which found a long-term drift rate of about 1 ppm
per 1000 hours when the die was self-heated to 88°C.

Other people have observed that if the LM399 isn’t heated
to 88°C, but just kept at room temp, the long-term drift rate
can be less than 1 ppm per 1000 hours. So should you heat
them up only an hour per month? Maybe so.

The V os of an op amp is almost trivial—and it is not trivial, nor
is it trivial to guess, what will happen on the next full temp cycle.
Sometimes an op amp will drift lor 2 pV. But other times it may
drift 3 or -4 pV after cycling around a full set of temperature tests.
The next time it goes around the cycle, it might drift 4 or -3 pV.
This is due to stress on the die. Can you predict this? I don’t think
so. So even an op amp requires some respect in its testing.

A computer can predict how much stress will be on the first
die, at various places, when it is packaged. It can predict how
much stress will be on the second die—and the third. It’s all the
same. So offsets and drift and hysteresis will be the same,
right? Not so. So much for computers. So much for CAD. I pre¬
fer to admit the reality of computer-hindered design.

These stresses apply also to band-gap references. They, too,
have drifts as they are temp-cycled and brought back to the
original temperature. I don’t know any circuits that aren’t more
stable if you just leave them alone. ©

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California

11 . 05.07 ELECTRONIC DESIGN

ED ONLINE 17436

• BOB PEASE | Contributing Editor

rap@galaxy.nsc.com

Bob's Mailbox

ob: In “What’s All This Capacitor Leakage Stuff, Anyhow?”
(March 29, 2007, ED Online 15116) you have a diagram
of a test circuit using the LMC662. This circuit is quite a
bit different from the “capacitor soakage” test circuit you
referenced on your Web site. (Afteryou get the capacitor charged
up, and after you get the soakage elements charged up (which
takes hours and days), the rate of change of V 0UT is caused by the
leakage, which you cannot really see until you have waited some
days. Any rate of change before that is mostly caused by soakage.
So in this experiment, I was trying to separate the soakage from
the leakage. After a few days, this did work right. Fora soakage
test, try charging up high first and then pulling it to ground. The
leakage is not involved./ra For the leakage circuit, I was curious
what the “ball hook” and the coax were for. What’s that all about?
(The ball hook lets me connect to several different capacitors in
sequence. It is insulated with cheap nylon. That isn't a very good
low-leakage insulator, but if I keep my body biased up to 8.8 V, it is
not too bad./rap) Why is the coax 12 in. long? (I’m guessing it
could be shorter but not a lot longer.) I figured the ball hook is the
same as what I call a “chicken stick.” We hams use them when
working on high-voltage circuits to make sure all the caps are for
sure discharged by grounding the chicken stick with a clamp to
chassis, then running the “ball-hook end” overall of the high-volt¬
age components. (Yeah, but when you use the chicken stick, do
you put it in series with 1 Q? Or 0.1 Q? How often does the 1-0.
resistor have to be replaced because you blew it out? I am not
working at low-ohms levels, but at 10 to 100 MMQ. Different
range, eh? By 13+ orders of magnitude. I hope that you have
learned that even your chicken stick can not discharge a big high-
voltage oil capacitor properly because it can recharge itself up to
hundreds of volts due to soakage if you wait a while. A big high-
voltage cap needs a resistor to discharge it for a long time to make
it stay at low voltage, not just for a second. Over near Livermore,
some movers were moving some big high-voltage capacitors. One
crate got busted, the shorting bar got knocked off the capacitor,
and a moving guy bumped into it and got killed—electrocuted—
even though the capacitor must have been shorted out for hours!
Dangerous things./rap) Here, you’re probably using it to prevent
distorting the experiment through human-body discharge. (Yeah, it
is an adequate insulator. It seems to have less than 2 pF of capaci¬
tance and 1 lousy MMQ. If I had a Teflon ball hook, that would be
nice. But the ordinary nylon one is okay. I don’t need perfection. I
can get good measurements if it is just good enough for a short
time, /rap) Just curious about the specifics with the coax. { Even
though I could not see it, I remember that the coax was to prevent
leakage into the + input. It had to be at least 12 in. long to reach all
the capacitors. If I had a bigger array of capacitors, I could have
used 2 or 3 ft of coax to reach them. No big deal, /rap) This looks

like a fun little project. Have you done any more work on it since?

• Jason A. Dugas

• Pease: No, but I will next week.

Bob: A year ago you had an article on the Sallen-Key filter (Sept.

28,2006, ED Online 13480). In that article, you mentioned that
using an op amp with higher input impedance will allow the use of
larger values of R1 and R2 and smaller, cheaper, and higher-quali¬
ty capacitors. What is the tradeoff of staying with smaller R1 and
R2 values to reduce Johnson noise and having to use X7Rcaps
instead of NPO caps? To use all NPO caps in my four-pole, two-op-
amp design, I must cross over the 1-MQ level for some R values.

• Bruce Allen

• Pease: What BW are you trying to handle? Low pass or high
pass? If you used 0.01-pF NPOs and 1MQ, you would be talking
16 Hz. A 1-MQ resistor doesn’t have much noise in a 16-Hz
bandwidth. It would have (125 nV X4) rms or 3 pV p-p. Most op
amps have more noise than that. These days, you can buy
NPOs bigger than 0.02 p F. Look in Digikey. You ought to build it
and try it both ways. Avoid X7Rs, which have long soakage tails.

Hi Bob: Nikola Tesla was an amazing man. He is now credited as
the inventor of the induction motor, wireless transmission, Tesla
coil, logic gates, Tesla turbines, lighting ioning gases, etc. (No argu¬
ment. His induction motors and several other items were quite bril¬
liant./rap^ I’m most interested in his claims of wireless transmis¬
sion of power. If anyone could have pulled this off, he was capable
of doing it. He invested a lot of his time and money in this endeav¬
or, only to fall through by being under-budgeted. (I think he was
over-ambitious with schemes that could not possibly work. Some
of his schemes were loony, and even Tesla couldn’t make them
work./rap I think we are just coming to grips with what he knew.
He knew the Earth’s capacitance and telluric currents, resonance,
distance of the ionosphere, the difference between transverse and
longitudinal waves, etc. What is your opinion on this? In general, do
you think it’s possible? Do you have any technical insights?

• Tony A. Wittic

• Pease: I am not 1/20 as smart as he was when he had his head
screwed on right, because I am not an electric field expert. I’m
just dumb. That’s okay. When he had nutball ideas, which he
often did later in life, he may have been dumber than me.

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

11 . 15.07 ELECTRONIC DESIGN

PEASE PORRIDGE

WHAT'S ALL THIS "BEST TRICK
CIRCUIT" STUFF, ANYHOW?

Bob Pease

CONTRIBUTING EDITOR

nee upon a time, op amps
didn’t swing very close to
the positive or negative
rails. Even a couple volts
away from the rails—that
was okay in the old days
of transistor-ized op amps.

Hey, that was a lot better than with
vacuum-tube op amps that wouldn’t
swing within 200 V of the rail. Some op
amps could do a little better, but cus¬
tomers never asked us for better 35
years ago.

The LM324 can swing pretty close to
ground (-Vs) if you have a pull-down
resistor—or close to +V S if you have a
pull-up resistor. But, of course, it has a
pretty big (ac) dead zone when driving
its output. Crude. Slow.

Many modern low-voltage op amps
have common-drain outputs. This is
often called a “rail-to-rail” output, even
though it won’t really go to the rail.
Martin Giles always razzes me (quite
properly) if I talk about these amplifiers
as “rail-to-rail.”

Widlar’s 1976 LM10 was the first to
have a common-collector “rail-to-rail”
output, swinging within a few millivolts
of each rail. Most CMOS op amps don’t
swing that close. If you have an op amp
running on “+5 V dc,” it surely can’t
swing to even + 4.9 V if its power supply
is only + 4.75 V (i.e., if the supply has a
5% tolerance).

So if you have an analog-to-digital
converter (ADC) with a 2.5-V reference,
and a 2.5-V full scale, an op amp run¬
ning on +4.75 V can easily swing up to

RRro

Fij ure 2

the + full scale at +2.50 V. But it
still can’t swing that close to
ground. Even for a 10-bit ADC,
you are going to lose a few least sig¬
nificant bits where the amplifier
can’t swing low enough (Fig. 1).

The output may swing within 12 or
22 mV of ground, but not closer.

The trick is to add a high-imped¬
ance diode from (a) pointing to (b),
replacing the hard wire. (1N4004, not
1N914). Even if the amplifier’s output
can’t swing within 100 mV of ground, it
doesn’t have to. To get the output to
within 0.1 mV of ground, the diode
leakage merely has to go down to 50 nA
at Vp = 0.1 V, even at your highest oper¬
ating temperature. Even at 95°C, you
can do this with a transistor’s C-B
diode. Try it.

The ~2k pull-down resistor in shunt of
the output and the diode in series with the
op amp’s output make the trick work. It
can drive the high-impedance input of an
ADC within a small number of microvolts
of ground—much better than a “rail-to-
rail” that can only get within a small num¬
ber of millivolts of ground.

So the circuit of Figure 1 can swing
close to ground nicely, but not close to
the positive rail. Is that what you need¬
ed, bunky? Why didn’t you say so? Try
Figure 2. Install one high-impedance
diode from (c) toward (d) and another
one anti-parallel.

The 74C14 has a lot of “gain,” but
not a lot of output offset—and very lit¬
tle power drain in this switching mode.
I have seen it drive an ADC within a
few microvolts of both rails.
One of the advantages is
that it won’t over-drive the
input of the ADC past the rails.
Good feature.

These circuits are not very fast,
and not low-distortion, and they
won’t drive much of a dc load. Do you
want to drive a dc load? If we add
another trick, you can drive a load. Go
to www.national.com/rap and click on
“rail-to-rail driver.” Have fun!

SCHOOL DAYS • When I was a kid
engineer of 18 at MIT, I took course
6.021 on Piecewise Linear Circuits with
graduate instructor Leonard Kleinrock. I
learned a lot from him.

Now, Prof. Kleinrock is a respected pro¬
fessor at UCLA and one of the inventors
of the networks leading to the formation
of the Internet. AND 49 YEARS OUT, I
AM STILL HAVING FUN DESIGNING
PIECEWISE LINEAR CIRCUITS! I hope
that Lenny is still having fun, too!

BEST NEW RECIPE • For an excellent
casserole, boil 1 lb white beans (-Great
Northern) in 8 cups H 2 0, 3 minutes; let
stand 2 hours, then simmer an hour.
Meanwhile, bake -2.5 lb of pork roast at
350°F. Cut into 1/3-in. cubes; discard
most chunks of fat. Scrub potatoes and
cut into 1/4-in. cubes to make 2 cups.
Add pork and potato to the beans. Add a
4-oz can of warm or hot chopped chilis,
to taste, and one (or two) (or 1/3, if you
are trying to not scare the kids) 4-oz cans
of hot chilis, chopped fine (3/16 in.). Add
salt and pepper to taste; simmer 1/2 hour.
RAP invented this Best Recipe of the year:
I had a hunch, and it worked out well. ©

Comments invited!

rap@galaxy.nsc.com, or
Mail Stop D2597A
National Semiconductor
P.0. Box 58090
Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961
and is Staff Scientist at National Semiconductor
Corp., Santa Clara, California.

ED ONLINE 17601

12 . 03.07 ELECTRONIC DESIGN

ED ONLINE 17697

• BOB PEASE | Contributing Editor

rap@galaxy.nsc.com

Bob's Mailbox

ob: While my full-time job is writing for Electronic Design, I
still teach part-time and work on an NSF grant that is
attempting to update the electronics curricula in commu¬
nity colleges. From my observations in my own college and
across the country, most curricula are out of date with what is
going on in the industry. (True, but not disastrously bad. We can’t
ask that education for techs or for EEs be really up to date. That
has almost always been impossible, /rap) There is too much
emphasis on BJTs and little or poor coverage of MOSFETs and ICs.
(Yeah, but you can get and buy and solder BJTs, and the circuits
will work. You can’t get MOSFET kit parts worth crap! And if you do,
they don’t work well. In most cases, even Spice works better than
making breadboards of MOSFETs. Of course, we know the excep¬
tions... If a kid has learned a little about MOSFETs and a lot about
BJTs, we can convert him over, /rap) I would love to get your opin¬
ion. Could you answer a few questions? First, what would you say
the percentage mix of BJT/MOSFET circuits is in ICs (or discretes)
for linear and digital? It must be close to 100% digital, but what
about linear? (I think it is about 45/55, but we work on weird proj¬
ects. Many people do not recognize that to make low-power cir¬
cuits, CMOS is not inherently low-power. It takes a lot of work. So,
we make micropower op amps using BJTs—and very fast ones, too.
When you go that fast, you have to trust Spice, somewhat, /rap)
Second, have you seen any engineering techs recently? And does
NSC employ them? (Yes and yes. I have interviewed some and lec¬
tured to some recently, /rap) These are the guys that help engi¬
neers with bread boarding, test, etc. I haven’t seen many in years,
although I used to be one. (I actually repaired Philbrick K2Ws and
related gear way back years ago.) What is your take on this? (Our
technicians can repair ANYTHING—yes, even K2Ws. (I still have
some and use them.)/rap) Third, how important is it for a tech to
know detailed BJT and MOSFET biasing, etc.? ( Generally, not. Bias¬
ing FETs is almost impossible because a good bias depends on the
match of the FETs, and that happens well on only one chip./rap)
Fourth, most of us working on the NSF grant think that a tech
needs more of a systems view today as opposed to a detailed cir¬
cuit-analysis background. Do you concur?

• Louis E. Frenzel, Communications/Test Editor

• Pease: I tend to agree. Often, a good technician must use and
understand op amps—and measuring equipment, DVMs, spec¬
trum analyzers, and automated test equipment. Circuit analysis
is a specialty, and even for engineers, this is challenging. Of
course, they should be aware of circuit analysis and bias setup.
Even 20 and 40 years ago, we did not demand them to be
experts at that. We’re now asking our senior techs to do more
and more analysis of data and to tell us when things look right—
and when things look suspicious or “funny.” They’re usually quite
good. Of course, we’ve had the luxury of several excellent techs

out of the College of San Mateo, and with just a few years of
mentoring, many of them have become circuit engineers.

Dear Bob: Your article “What’s All This Input Impedance Stuff, Any¬
how?” (Sept. 7, 2004, ED Online 8576) describes a single op-amp
differential amplifier circuit with a gain of 100; resistor pairs of R1
= lk and R2 = 100k; and front-end buffers ignored. I do not argue
about circuit gains of voltages. They seem to be okay. But the
impedances are not. Various sources (like NS’s Linear Application
Flandbook: AN-20, AN-29, etc.) state that the input impedance of
the circuit at the inverting input is equal to Rl. I do not accept this
unconditional statement! When inputs are equal and in opposite
phase, as in an ideal case, the input impedance is actually 500 Q.
(Ah, but that is a special condition!/rap) In your article, the imped¬
ance is claimed to be Rl = lk (according to the handbook state¬
ment), which is incorrect. When inputs are driven to 10 V dc each,
the impedances are 101k on both sides, just as in your article. I
agree. But this 101k impedance is not equal to Rl = lk. If gain = 1
(all four resistors equal), then impedance is 2/3 * R. I argued
about this with Dr. Michael Ellis a few years ago, and after some
calculations and simulations, we agreed. (First, I was wrong when
believing the unconditional statement above.) But when the input
voltages are not equal, the situation changes further. For example,
if positive input is 100 times negative input and in opposite phase
(gain = 1 circuit), then the impedance is only 1/50 * R. You may
simulate various conditions as I did and find out that the imped¬
ance varies vastly. (Well, if you pick the right “ special conditions,”
anything can happen!/rap) Obviously, we must come to the con¬
clusion that the impedance at the inverting input node is not equal
to 1, neither constant, but depends on magnitudes and phases of
input voltages and is therefore largely variable. In general, when
inputs are not correlated with amplitude or phase (random input or
noise), one can not guess the impedance. Do you agree?

• Eero-Pekka Mand

• Pease: I tend to agree. But (a) there is nothing simple about
this, and (b) you have waited three years to comment! So (c),
shall we consider the case where the input is a transformer
winding, not center-tapped? Or(d) cap-coupled? If the caps are
big enough, it may still work. But you may need 100X bigger
capacitors on the negative input. Let’s discuss. This might lead
to “ What’s All This Z In Stuff, Anyhow? (Revisited).”

Comments invited! rap@galaxy.nsc.com —or:

Mail Stop D2597A, National Semiconductor
P.O. Box 58090, Santa Clara, CA 95052-8090

BOB PEASE obtained a BSEE from MIT in 1961 and is Staff Scientist at
National Semiconductor Corp., Santa Clara, California.

12.13.07 ELECTRONIC DESIGN

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Texts

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Bob_Pease_Lab_Notes_Part_6"

See other formats